Understandable for beginners . . . interesting for experts
May 1978 / Issue #17/ $2.00 /DM 7,50 / Sfr 8, 10 / Ffr 16,0/ UK £2
Ralph Wells
22
Bob Buckman
32
J. Tom Badgett
36
Robert Baker
42
Dan Stogdill
44
Peter Stark
48
George Young
54
John Blankenship
60
Sheila Clarke
64
Dr. Lance A. Leventhal
68
Howie DiBlasi
76
Stephen Gibson
78
Dr. Adam Osborne
84
Ken Barbier
90
Richard Roth
94
Thomas E. Doyle
100
Mike Kop
104
Dave Waterman, Dave Lien
110
Glen Charnock
112
PET’S First Report Card ... an objective evaluation
Scope Power! ... a review of Tektronix’s Model 922
Trials and Tribulations . . . one businessman’s micro blues
Writing Diagnostic Routines . . . while your machine is running
Experiments in Software . . . serial to parallel conversion
Computer Math Primer . . . beginner’s introduction to number systems
Kilobaud Klassroom ... No. 10: Bus Traffic Control
Expand Your KIM . . . Part 5: A/D interfacing (for joysticks!)
What’s Happening with the IBM Selectric?
The Top-Down Approach . . . with some practical examples
The North star Floppy System ... an 11-year-old can build it!
A Simple Mailing System ... a money-making time-saver
Number Crunching: Two Hardware Solutions
Money Manipulations . . . keep ahead of those cash-flow problems
Strings and Things . . . BASIC conversion techniques
5 Minutes or 5 Hours? . . . sorting techniques compared
Do-It-Yourself Time-sharing . . . it’s easier than you think
Cassette Recorder Disaster: Ground Loops
A Different Search Technique . . . don’t just try it— benchmark it
Publisher’s Remarks— 4, Editor’s Remarks— 6, Around the Industry— 6, TRS-80 Forum— 8,
Legal/Business Forum— 12, KB Club Calendar— 13, New Products— 14, Books— 16, Letters— 17,
Kilobaud Classified — 114, Contest! — 114, Calendar — 116
READY lor BUSIHESS
We've got it all together— the cost effectiveness and reliability of our 6800 computer system
with a high capacity 1.2 megabyte floppy disk system. . . PLUS— an outstanding new DOS
and file management system.
1 MEGABYTE DISK SYSTEM
DMAF1 introduces a new level of capability to
small computer systems. This disk system fea-
tures two standard size floppy disk drives using
the new double sided disk and two heads per
drive. Usable storage space of over 600 kilobytes
per drive, giving a total of over 1 .0 megabyte of
storage on line at all times. Ideal for small busi-
ness applications, or for personal "super” sys-
tems.
DMA CONTROLLER
The controller occupies one main memory slot
in an SS-50 bus and uses the Motorola MC-6844
DMA controller. The combination of a DMA
type controller and double sided disks give the
system speed of data transfer unobtainable with
smaller drives.
OPERATING SYSTEM
To compliment this outstanding hardware we
are supplying equally superior software. The
disk operating system and file management sys-
tem is called FLEX. It is one of the most flex-
ible and complete DOS's available for small sys-
tems, but just as important; it is easy to use.
No one can match the variety of compatible
peripherals offered by Southwest Technical
Products for the SS-50 bus and the 6800 com-
puter system. Now more than ever there is no
reason to settle for less.
DMAF1 Disk System (assembled) $2,095.00
DMAF1 Disk System (kit) $2,000.00
68/2 Computer with 40K of memory (assembled) $1,195.00
SOUTHWEST TECHNICAL PRODUCTS CORPORATION
216 W. RHAPSODY
SAN ANTONIO, TEXAS 78216
For details write or call today.
0 (TO U 0
PUBLISHER’S
REMARKS
The snow required by nearby ski areas tends to obscure buildings in
New Hampshire. Here's Infotecs’ plant, somewhere under the snow.
The Digital Group
As a continuing part of my
plan to personally visit as many
of the major firms in the micro-
computing industry as possible, I
recently stopped by The Digital
Group facility in Denver.
They’ve recently moved to new
and larger quarters, a move I
envy. Success is almost as diffi-
cult to cope with as failure — it’s
just a lot less painful. TDG has
been going through growth prob-
lems — and they are many — for-
tunately, with some success.
One of the reasons for some of
the horror stories about early
Mits service was the over-success
of the 8800 Altair. It is almost im-
possible to go through cata-
strophic growth without ex-
periencing severe traumas. The
“science” of selecting personnel
is not yet perfect. It takes weeks
or months to find out whether the
peg is round or square, so the end
result is that you have to train two
or three people for each job — an
expensive and time-consuming
situation — and all this is happen-
ing under great pressure from
frustrated customers.
You can imagine the things
that can go wrong as you try to
expand from a very small group
— I think Mits had about 15 peo-
ple when they announced the Al-
tair — to ten and 20 times your
original number. Everything that
can go wrong will . . . repeatedly.
TDG has been going through
this catastrophic growth and
seems to be emerging reasonably
Wayne Green
well. Despite the pressures on
them to deliver systems already
being advertised, they’ve been
doing their homework and have
new systems ready for showing.
That’s even more remarkable
when you consider that they’ve
just moved to a new building.
The new building should hold
them for a while. We’re so
packed in here at Kilobaud , with
about 70 people jammed into an
old 40-room house, that those
modern, spacious offices— some
not even being used yet — caused
some slight pangs of jealousy.
The Infotecs Bombshell
While most of the manufac-
turers in the microcomputer field
have been concentrating on de-
signing and selling hardware,
with just enough software to get
the hardware to sell, at least one
firm has been taking a different
tack.
Three years ago I looked over
the amount of money 73 Maga-
zine was spending on outside data
processing and was amazed to
find it was around $2000 a
month. For that amount of
money we could have an in-house
system that could do what was
needed . . . plus much more. I
sent out word to the computer
world that I needed a computer.
The salesmen flocked in, each
with wondrous descriptions of
how great his system would be
and with conspiratorial referenc-
es to competing systems and their
virtually fatal flaws. The main
problem for me was that I
couldn’t understand much of
what the salesmen were saying.
They spoke computerese and ap-
parently had not even a vestigial
recollection of English.
In my efforts to learn comput-
erese, I stumbled through com-
puter hobby newsletters and
microcomputers. I fell for micro-
computers, hook, Mits line and
sinker. The newsletters were OK,
but there really should have been
a magazine ... so I started call-
ing the editors of the computer-
club newsletters to see if I could
find someone who might be will-
ing to take a gamble. Hal Cham-
berlin wasn’t interested. Neither
was Hal Singer, nor Bob Al-
brecht. I finally got down to a
chap named Helmers who had a
newsletter with 300 subscribers
and he said, “OK, let’s try it.”
On the same day, as I recall, I
came up with the name for the
new magazine: Byte.
During the five weeks after the
decisions to get cracking on Byte ,
I rounded up authors and mailing
lists from manufacturers, wrote
subscription letters, wrote to all
of the clubs, put out newsletters,
etc. It was during this time that
Cal Holt stopped by to try and in-
terest me in buying a PDP8/A to
take care of the 73 /Byte subscrip-
tion list, bookkeeping, orders,
etc. While Cal was trying to sell
me on buying a PDP8/A with his
programming, I was busy selling
him on what was happening in
microcomputers and how I en-
visioned the future of this new
industry.
It took us just five weeks to get
the first issue of Byte on the
press. This was five weeks of day-
and-night work for me, but the
first issue ran 15,000 copies in-
stead of my original hope of
2000. The magazine got off to a
good start . . . and so did the
microcomputer industry. In
November I managed to lose the
magazine ... a story I hope the
lawyers will eventually allow me
to publish. There are too many
lawsuits going now, so you’ll
have to wait that one out.
Now, flash ahead about 2Vi
years to January 1978. Cal Holt
started calling and leaving mes-
sages for me to get up to see his
plant in Manchester (NH). What
with CES in Vegas and a ham in-
dustry conference in Aspen, plus
a long siege of the flu in January,
it took me a while. When I finally
made it to Infotecs it was worth
the trip.
Though I haven’t been making
a big deal out of it in Kilobaud , I
have been discussing the change
in the microcomputer market in
our Kilobaud Newsletter, which
goes to the industry. The fact is
that hobby growth has essentially
stopped, and most of the growth
in the industry has been in sales to
small business. Since I had pre-
dicted this right from the begin-
ning, this has not come as any
surprise to me.
Being as trapped by the com-
puter hobby as anyone, I have no
intention of putting it down. But
my rational has been this: The
The CRT and CPU units are tested here as they are
completed.
Here the complete systems are checked out and run
in. The whole Infotecs microcomputer system is
thus put together and checked out in a relatively
small plant in Manchester, New Hampshire. How
big a plant will they need a year from now?
4
The CR T, printer and disk that make up the Infotecs complete system.
The 6100 CPU and memory are in the disk unit.
computer hobby is a very de-
manding one, requiring a lot of
work and expense for the hobby-
ist who is actually going to under-
stand computers and work with
them. It is, in computerese, a
nontrivial hobby.
I felt that this demand on the
individual would be a limiting
factor. How many people could
we find who would take the enor-
mous amount of effort required
to become serious hobbyists?
This had to be a limiting factor,
whether it came at 100,000 hob-
byists or 1,000,000. My predic-
tions were more in the 100,000
range.
Why did I put the figure that
low? Well, I know how easy it is
to get a ham license as compared
to understanding computers, and
I know how the number of hams
has been limited by the effort re-
quired to learn the theory and the
code, both trivial compared to
computers. It seemed like a rea-
sonable yardstick.
Talks with people at computer
stores during the last year have
convinced me that my predictions
were not far off. Most stores are
reporting that sales to hobbyists
have not changed seriously dur-
ing the last year, but that sales to
business have come along from
nothing to about four or five
times those to hobbyists. Those
stores that are particularly hobby
oriented have been reporting
about equal sales between the two
factions. You can quickly spot a
hobby store when you walk in
. . . you’re ignored unless you
are a hobbyist. I don’t care if you
have $10,000 burning a hole in
your pocket, they will fawn all
over a kid playing Star Trek and
pointedly be deaf to any ques-
tions you may have.
One of the major problems
stores face in selling systems to
small businesses is the lack of
business programs. Few business-
men want to spend $12,000 or so
getting the hardware, only to
have to sit down and write their
own programs. Some stores have
been busy writing programs in
order to facilitate sales, but this is
awfully expensive.
Imagine my surprise and de-
light to visit Infotecs and find
that they had developed a com-
plete microcomputer system of
their own, including about the
most comprehensive fuel-oil-
dealer package I could imagine.
The system is based on the Intersil
6100 chip, which emulates the
PDP8 . . . thus giving Cal and
his programming staff a good
headstart by virtue of their work
with the PDP8. The fuel-oil-deal-
er package was mostly written by
Infotecs president Ed Tolson,
and you really have to see it to
believe how complete it is . . .
right down to providing a print-
out of a customer list showing the
overall profit made on each
customer.
The program keeps track of
each oil truck and its service,
sales, route, driver, etc. It bills
and sends statements to the cus-
tomers. It is most complete. Vir-
tually every oil dealer who has
seen the system has signed up for
one . . . including two in Peter-
borough.
Infotecs has another program
they are just releasing. This one is
for accountants; it, too, is most
comprehensive. Infotecs buys the
printer, the keyboard and moni-
tor, and makes the microcomput-
er board themselves. The whole
system sells to the customer
through computer stores or other
dealers for about $18,000, which
comes to under $350 per month
on a lease. At that price, no oil
dealer can afford not to get one.
I think Infotecs could sell
thousands of their oil-dealer sys-
tems if they could make them fast
enough and find dealers with
enough backing to be able to han-
dle the business. And once they
get going on promotion of their
accounting system, who knows
what could happen?
Of particular note is the disk
system used by Infotecs. They’re
using a PerSci disk with their own
operating system. The dual disk
holds almost two megabytes . . .
that’s right, 946,176 bytes per
disk. How can they do this? One
of the tricks is to split the 12-bit
words into two characters of six
bits each. They have to forego
lowercase to do this, but they can
add lowercase if they want to pro-
vide a word-processing system la-
ter by changing to 12-bit char-
acters.
The CRT has 24 80-character
lines, one of the largest video dis-
plays in the business. The printer
runs 132-character lines, dot ma-
trix, at 125 lines per minute.
I talked with some fuel-oil
dealers to see how they felt about
the Infotecs system. Those al-
ready using it are very enthusi-
astic. The owner himself is able to
do the data input if he wants and
thus keep control over his busi-
ness. The system checks billings
against gallons delivered and
warns the operator if things don’t
add up. It also handles such side-
lines as diesel oil, gas sales, pro-
pane sales, furnace cleaning, etc.
The dealers explained that the
Infotecs system is much cheaper
(continued on page 20)
Reader Responsibility
One of your responsibilities, as a reader of Kilobaud, is to aid
and abet the increasing of circulation and advertising, both of
which will bring you the same benefit: a larger and even better
magazine. You can help by encouraging your friends to sub-
scribe to Kilobaud. Remember that subscriptions are guaran-
teed— money back if not delighted, so no one can lose. You can
also help by tearing out one of the cards just inside the back
cover and circling the replies you’d like to see: catalogues, spec
sheets, etc. Advertisers put a lot of trust in these reader re-
quests for information. To make it even more worth your while
to send in the card, a drawing will be held each month and the
winner will get a lifetime subscription to Kilobaud !
kilobaud
J
7 S
Publisher
Wayne Green
Executive Vice President
Sherry Smythe
Editor
John Craig
Managing Editor
John Barry
Editorial Assistants
Dennis Brisson
Susan Gross
Administrative Assistant
Dotty Gibson
Production Department
Lynn Panciera-Fraser
Craig Brown
Gayle Cabana
Robert Drew
Michael Murphy
Weston Parker
Noel R. Self
Robin M. Sloan
Typesetting
Barbara J. Latti
Pauline Halvonik
Jennifer Johansson
Marie Walz
Photography
Bill Heydolph
Tedd Cluff
Associate Editors
Don Alexander
Tim Barry
Sheila Clarke
Rich Didday
Phil Feldman
Tom Rugg
Peter Stark
Bookkeeper
Knud E. M. Keller
Marketing
Sherry Smythe
Cynthia Gray
Circulation
Barbara Block
Fran Dillon
Florence Goldman
Rhonda Ramsey
Receptionist
Doni-Anne Jarvis
Computer Data Control
Judy Waterman
Judy Brumaghim
Linda Cate
Mary Kinzel
Computer Programming
Ron Cooke
Richard Dykema
Steven Lionel
Printing
Michael Potter
William Cering
Dwight Perry
Mail room
Theresa Toussaint
Bill Barry
Sue Chandler
Ethan Perry
Advertising
Heidi Kulish
Marcia Stone
lla K. Witty
European Distributor
Monika Nedela
Australian Distributor
Katherine Thirkell
UK Distributor
L P Enterprises
Kilobaud is published monthly by 1001001, Inc.,
Peterborough NH 03458. Subscription rates in the
U.S. ana Canada are $15 lor one year and $36 for
three years. In Europe: Kilobaud erscheint mon-
tatlich bei Fachzeitschriftenvertrieb Monika Nedela.
7778 Markdorf. Markstr. 3. Abonnement DM 70 - plus
Porto 7.20 + gesetzl. MWST; Sfr. 81. plus Porto 7.20.
Australia: For subscriptions write— Katherine
Thirkell, Sontron Instruments, 17 Arawatta St.,
Carnegie, Vic. 3163 Australia. UK: £20 ster-
ling/year— LP Enterprises, 313 Kingston Rd., Ilford,
Essex. IG1 1PJ, England. Please write for other
foreign rates. Second class postage paid at Peter-
borough NH 03468 and at additional mailing ufflces.
Publication No. 346690. Phone: 603 924-3873 Entire
contents copyright 1978 by 1001001. Inc. INCLUDE
OLD ADDRESS AND ZIP CODE WITH ADORESS
CHANGE NOTIFICATION.
5
EDITOR’S
REMARKS
John Craig
Finally: An Affordable
Computer-Portrait System
You may have noticed my en-
thusiasm for computer-portrait
systems in previous comments
I’ve made about them. I’ve al-
ways felt that the “hobbyist”
community could, and would,
come up with something that
would sell for considerably less
than the $20,000 to $25,000 such
systems are currently going for.
Well, it finally happened . . . and
the Micro Works in Del Mar CA
is the company that did it. The
photograph in the January issue
of the Mona Lisa being generated
on a Malibu Design printer
caught their eye . . . and that’s
the printer they selected for the
system. Photo 1 will provide a
glimpse into what the “Micro
Workers” are up to. We’ll see if
we can get the full details, along
with some spectacular portraits,
in next month’s issue.
Cybervision Update
Last month I commented on
Montgomery Ward’s new entry
in the home-computer field . . .
and said I’d try to get more de-
tails for this month. I’m happy to
report that I not only got the de-
tails; but it looks like we’re going
to have a review of the Cybervi-
sion in an upcoming issue of
Kilobaud.
I really should point out that it
wasn’t the hardware that turned
me on about this system. After
all, it sells for $400 and only con-
sists of a box (with a reset switch),
a cassette recorder and two inex-
pensive touch-pad keyboards. No
. . . what excited me was that it
was the first home system to be
introduced to the American pub-
lic through one of the major mail-
order catalogues, and I feel that’s
significant. Also, it occurred to
me that this might be an ideal sys-
tem for the hobbyist who hasn’t
made the plunge. The Cyber-
vision’s price is low; software is
available that the family can put
to use right away (for education
and entertainment); expanding it
in several directions is fun and
challenging.
The system is being manufac-
tured for Montgomery Ward by
Associated Sales in Columbus
OH. After talking to Associated
Sales’ chief engineer, Jim Mc-
Connel, and Joe Miller, their sys-
tems programmer, my enthusi-
asm has certainly increased. Jim
filled me in on some very in-
teresting hardware details, not
the least of which is that the
Cybervision is built around the
RCA 1802 microprocessor. My
first reaction was, why? He
pointed out that because the 1802
is a CMOS chip it not only has
low power consumption (and
other features) but also generates
very clean waveforms, which re-
sults in negligible generation of
Photo 1. The Micro Works system; TV camera on the left , SWTP
system; Malibu Design Group printer.
AROUND
THE INDUSTRY
John Craig
The Noval 760: Here it Comes!
Do you remember the Noval
760? It first appeared in a full-
color ad in the June 1977 issue of
Byte . . . and captured the imag-
ination of the entire industry. The
system is unique, and one of the
big reasons is its “packaging.”
Since a picture is worth a thou-
sand words (I just made that up),
I won’t strain my typewriter try-
ing to describe the beautiful desk
the system comes in— just take a
look a Photos 1 and 2.
Actually, that desk is the rea-
son why you haven’t been hearing
too much of the Noval 760 in re-
cent months. They’ve gone
through three suppliers in an at-
tempt to get it manufactured to
their specs. The hardware and
software was debugged long ago,
but the desk has kept them from
accepting orders and advertising
the system.
The additional efforts have
certainly been worthwhile. You
know, that computer doesn’t just
pop up out of that desk. Instead,
it rises up slowly and gently and
brings forth a lump in your throat
much as the playing of the Na-
radio frequencies. Because of
this, getting FCC approval was
much easier . . . and the unit
doesn’t require shielding.
Joe Miller noted a really fasci-
nating item regarding the soft-
ware. The system only comes
with 4K of memory; but because
of the use of overlay techniques,
programs larger than 4K are run
with ease. It takes about eight sec-
onds to load a 2K program seg-
ment (via their 2000 baud cassette
interface), and such transfers are
hidden, or masked, from the user
by voice prompts or responses
taking place (on the second chan-
nel of the recorder) during the
transfer.
The Cybervision has other neat
features, but let’s save them for
the review. A refreshing com-
ment from Associated Sales was
that they don’t consider it a threat
to have other people and compa-
nies building add-ons and periph-
erals and generating software for
their system. As a matter of fact,
they’re going to be making details
of their bus and interfacing infor-
mation available for that upcom-
ing article— and their own soft-
ware development interpreter will
be made available in the future.
(By the way . . . Tom Pittman
will produce a Tiny BASIC for
the system in the near future.)
“The Colonel” Goes AWOL
Notice: Norman Henry Hunt
(alias Colonel David Wintrop)
has pulled off another con. Hunt
cut through a fence at Chino
(CA) State Prison, where he was
doing time for grand theft and
fraud (see Editors Remarks, Nos.
10-13), and escaped. He may be
setting up shop again. Next
month, we’ll have a photo of
Hunt plus information from the
detective who arrested him.
tional Anthem! (Maybe that only
happens with red-blooded com-
puter nuts . . . and not everyone.)
The Gremlin Connection
With the problem Noval has
been experiencing with the desk,
it’s not too farfetched to imagine
they might not have survived.
However, they are a sister com-
pany of Gremlin Industries, and
the necessary support has been,
and will continue to be, with them
(thank goodness for big sisters!).
Gremlin Industries is one of the
leading manufacturers of elec-
tronic arcade games; and the
veloped some outstanding educa-
tional games for grades 2 through
6. The games make extensive use
of graphics and generate a lot of
initiative and competition when
played by two students . . . since
the responses are timed and the
person with the fastest (and cor-
rect) answer wins.
A “People” Company
Photo 1. Is it just another nice-looking desk? No, it's a Noval home
system ... in its “sleeping” position.
Noval 760 is an outgrowth of
their efforts in developing
microprocessor-based video
games.
A few years ago, someone
popped up and said, “Hey, why
don’t we put together a computer
system with all this know-how we
have?” One of the fantastic
bonuses you get with the pur-
chase of a Noval system is the
availability of their video arcade
games to run on the machine.
(That’s right, they’re the same
games you have to pay a quarter
to play down in your local pub or
arcade!) Photo 3 illustrates the
system’s graphic capabilities in
one of Noval’s most popular
games, Depth Charge.
The games, and graphics capa-
bilities, offered with the 760 are
an important part of the overall
system approach. The system is a
home computer , aimed at the
hobbyist as well as the lay user.
Since most home-systems buyers
have entertainment applications
in mind, you can appreciate
Noval’s emphasis in this area.
And . . . you’ll be hard pressed
to find interactive video games
such as theirs on any other
systems.
I feel that just as much emphasis
should be placed on educational
programs as games — combining
them is even better— so I was
quite pleased to discover Noval
has done just that. They’re in-
volved with the San Diego School
District in a research program
called Telemath, and they’ve de-
When I arrived at Noval/
Gremlin, the first thing to greet
my eyes was an enormous 40-foot
banner spread across the front of
the building. “Happy Birthday,
Lonnie Pogue,” it proclaimed. 1
thought that was neat. 1 don’t
think you can appreciate the sig-
nificance of the sign, and how it
reflects the attitude of the com-
pany, until you stop and ask
yourself if something like that
could happen where you work.
Probably not, right?
They do some other “strange”
things at Noval that you won’t
find at most companies. For ex-
ample, they have such a dedicated
group that it is not uncommon
for individuals to become so en-
thusiastic and engrossed in what
they’re doing that they contribute
a lot of personal time to comple-
tion of projects. This may not be
too hard to understand when you
consider, after all, that one of
their primary products is games.
And the fun and games are cer-
tainly in evidence. The Gremlin
arcade games are set up every-
where . . . and available for the
employees to play with in their
off time, as you can see in Photos
4, 5 and 6.
This “people-oriented” atti-
tude is reflected in the design of
the Noval 760, also. The system
software was designed with the
average hobbyist in mind . . .
not the professional program-
Photo 6. In the development lab
new games must be tested ... in
the name of engineering research.
Photo 3. Depth Charge, one of the more popular
arcade games available for the Noval 760.
Photo 4. The employee lounge at Gremlin.
Photo 5. It's really rough . . . games everywhere!
Photo 2. The 760 . . . “awakened,” and ready for
action.
7
Photo 7. The heart of the 760 system.
mer. It’s forgiving, but not slop-
py; it’s sophisticated, but not
complicated; and it’s capable of
doing serious home software de-
velopment — within certain limi-
tations (due to a cassette, rather
than disk operating system).
The 760 Ingredients
The original design was an
8080-based system, which has
since been upgraded to a Z-80.
Photo 7 shows the rear of the 760
opened up to expose the system
board (the whole ball of wax is
right there). In the foreground
(left front) is the fully socketed
32K of memory that comes with
the system. To the left, plugged
into the three-connector back-
plane, is the 760 BASIC in
PROM. Additional RAM or
PROM memory segments can be
plugged into that bus. On the far
right, in the back, are the various
interface cards; the video graph-
ics circuit is situated in front of
them on the main board. So much
for the main board — now let’s see
what goes with it.
The peripherals mounted in the
console consist of a 2500 bps digi-
tal Phi-Deck cassette unit, a 12
inch b & w monitor (color op-
tional), a 32-column dot-matrix
printer and a full ASCII key-
board. I don’t know if there are
future plans to add another cas-
sette drive, but a 760 with dual
minifloppies is on the drawing
board. Larger printers can also be
ordered, at additional cost.
Prices for the 760 start at $3385
. . . and include all the hardware
I’ve mentioned.
The software provided with the
system consists of a monitor pro-
gram, a text editor and an assem-
bler. I found all three easy to use
. . . and quite adequate for soft-
ware development. Just to give
you another example of the hu-
man-oriented approach in the 760
design, you can use an illegal in-
struction mnemonic when assem-
bling a program. The assembler
will flag you that it is an error;
but if you insist on leaving it in, it
will simply be changed to a NOP.
Another nice feature is the listing
of the Editor commands on the
screen when you call the Editor
up for use.
BASIC is extra. The additional
cost (under $300) will get you
Noval BASIC or Noval Extended
BASIC installed in PROM . . .
with an added bonus of freeing
up the RAM memory it would
normally occupy. Their BASIC
allows full interaction with
assembly-language programs
(loading and executing), output
and input from individual I/O
ports and (here’s the big plus)
easy-to-use color and b & w
graphics commands. There aren’t
any string-handling functions in
the “standard” version, but they
are available in the extended.
(continued on page 20)
TRS-80
FORUM]
Dave Lien
How to Begin
Something like This?
By way of introduction, I’m
the culprit who wrote the operat-
ing manual for the TRS-80. Hav-
ing been part of the project from
its earliest days, I have some
small acquaintance with the sys-
tem. Perhaps in a future Forum
we can wax nostalgic and talk
about entertaining and humorous
events that are part of the proj-
ect’s history.
I work as a college adminis-
trator. Hobbies include comput-
ers (obviously), writing (maybe
not so obviously), ham radio, fly-
ing, sailing; and I keep busy con-
sulting. I have no special ax to
grind with or for Radio Shack.
The important person in this
Forum is you. I’d really rather be
off running a computer, but John
Craig has a mean hammerlock.
And Away We Go . . .
What do you want to see in the
TRS-80 Forum? It’s very impor-
tant that you let me know. Let’s
try the following format and see
what happens.
Flashes from Fort Worth
You aren’t alone out there!
TRS-80 sales are very strong. It’s
quite a surprise that about a third
of the Level I units are being
ordered with 16K of RAM. That
was a real sleeper.
Read it and weep! Field failure
rate is running a very low two per-
cent — not counting, of all things,
power-supply failures. Thought
that kind of problem was solved
around 1932. Seems some well-
meaning soul substituted fast-
blow for slow-blow fuses in one
batch of the sealed units. Any-
way, the problem’s supposed to
be solved now, and units being
delivered have the right fuse.
Much worse things could have
gone wrong.
All “factory type” inquiries re-
garding the TRS-80 should be ad-
dressed directly to Hugh Mathi-
as, customer service manager.
You can call him directly at (817)
390-3583. Hugh and his crew do
yeoman service, and I think the
enlightened attitude that RS is
taking towards customer service
is going to go a long way towards
making the TRS-80 the world’s
all-time best-selling computer.
Best User Program
of the Quarter
There is a classic computer
printout from the “heavy
frames” that goes back many
years: Snoopy shaking his fist at
the Red Baron. A microcomputer
version of that program simply
has to be added to each of our
software libraries.
Tom Kasper of San Diego,
one of the first TRS-80 owners,
gets software honors this time
around for his version of Snoopy
in RS shorthand with full TRS-80
graphics. Though we won’t com-
ment on his programming style
(and we’re not going to be too
fussy about that sort of thing with
any contributors unless some ma-
jor improvements are manda-
tory), we think his final product
on the screen is just superb ! Beau-
tiful job, Tom!
Why not submit your own fa-
vorite original program (on any
subject)? (Don’t worry about not
being a professional pro-
grammer.) Your fellow TRS-80
owners are hungry for good pro-
grams, and will probably “mas-
sage” it a bit to suit their own
fancy, anyway. Never did see two
programmers who completely
agreed on how best to write a
program.
We’ll be able to publish short
routines in the Forum, but, gen-
erally speaking, lengthy pro-
grams (such as Tom Kasper’s)
should be submitted as articles.
Be sure to send your contribution
(to Peterborough) on cassette
tape; a hard copy would also be
helpful. Record it several times
on the tape and include a tape
mailer and two first-class stamps
if you want the cassette back.
Computer-Scanning the Future
Level II BASIC is not far away.
It is nearly finalized at this writ-
ing and looks very good. While
incorporating the most valuable
of the standard Microsoft fea-
tures, it also retains most of the
Level I features.
The cassette Data flow rate has
been nearly doubled, and the
built-in editor is excellent — simi-
lar to the SOS editor used on big
DEC machines. Several printers
are available but require use of
the extra card box for interfacing.
Level III (for use with floppy
disks) is close behind, and is a
simple upgrade from Level II.
Matter of fact, most of what’s
needed for Level III is already in-
cluded in the Level II interpreter.
More on the future as we get
closer to it.
Feedback from the Field
30000 END
30370 REM *SIN* INPUT X IN DEGREES. OUTPUT Y.
30380 IF X>360 T. X = X/360 : X = (X-INT(X))*360
30390 IF X>90T.X = X/90:X = INT(X* 1000 + .5)/1000: Y = INT(X):G.30394
30392 G. 30400
30394 X = (X-Y)*90:ONYG. 30410, 30420, 30430
30400 X = X/57. 29578 : G.30440
30405 G.30440
30410 X = 90-X : G.30400
30420 X = -X: G.30400
30430 X = X-90 ; G.30400
30440 Y = X-X*X*X/6 + X*X*X*X*X/120-X*X*X*X*X*X*X/5040
30450 Y = Y + X*X*X*X*X*X*X*X*X/362880 : RET.
Fig. 1. Dave Waterman’s solution.
owners: little software or hard-
ware tricks you’ve discovered,
questions of common interest,
complaints or just plain com-
ments. The sooner you respond,
the sooner we can share your
feedback with others. Please
write (don’t phone): 8662 Dent
Drive, San Diego CA 92119.
Dave Waterman of Alpine CA
has contributed this valuable
feedback. It is a modification to
the SINE subroutine found on
page 218 of the TRS-80 User’s
Manual. To confirm that a prob-
lem really exists, enter the sub-
routine as printed plus the follow-
ing lines:
10 N = 179.95
20 X = N
30 GOSUB 30380
40 PRINT N,
50 N = N + .001
60 GOTO 20
and RUN
The crash of the subroutine is
due to the extremely tiny numbers
encountered as the value of the
SINE of 1 80 ° (and multiples of it)
is approached. Dave’s solution is
shown in Fig. 1 and seems to
solve the problem without intro-
ducing any new ones. Why not let
the other readers of the Forum
know how it works for you?
Meanwhile, we’ll see how Dave’s
new subroutine works in a really
super Level I biorhythm program
he is working on, which should
appear soon.
All things considered, the first
printing of the user’s manual
came out pretty well, but even
small typos cause great distress in
a computer program. By far the
greatest initiator of cards and let-
ters is the typo in the Combined
Function and ROM Test on page
227. Line 330 should read:
330.F.X = OTOA : A(X) = Q : N.X
It seems that thousands of
users thought they had bad ma-
chines because the factory test
program said so. Well— it shows
that users really are reading the
manual.
More on these and other things
in a later Forum.
3 REM SNOOPY BY T.N. KASPER, 5 JANUARY
5 CLS
10 F.A=14T022*S. (A,0) iN.A
20 F.B=12T015:S. (B, 1 ) sN.B
30 F.C=21T024:S. (C, 1 ) *N.C
40 F.C=40T045:S. (D, 1 ) :N.D
50 F.E=10TO12*S. ( E , 2 ) :N.E
60 F.F=21T025«S. (F,2) iN.F
70 F.G = 36T039: S. (G, 2 ) iN.G
1978 .
Tom Kasper’s Snoopy listing,
(continued on page 20)
That’s you! Since this is the
first Forum, there is only one
item of major interest from “the
field.” Only you can solve that
little dilemma.
Of special value will be items of
broad interest to other TRS-80
80 F . H= 4 5 TO 4 7 iS. ( H , 2 ) :N.H
90 F. I =9TO 1 0 : S . (1,3) »N. I
100 S. ( 12 , 3 )
110 F. J=23T035:S. ( J, 3 ) «N. J
120 F.K=45T049*S. (K,3) iN.K
9
INTERTEC’s
VIDEO DISPLAY TERMINAL
Something for Nothing!
Intertec's new INTERTUBE™ video ter-
minal is really something! But most amazing
is that it costs you almost nothing! The
INTERTUBE's $784* price tag combined
with our innovative microprocessor design
sets new standards in the CRT industry.
The new INTERTUBE obsoletes dumb ter-
minals and out performs the smart ones.
Every INTERTUBE boasts such features as
a full 24 line by 80 character display, 128
upper and lower case ASCII characters, re-
verse video, complete cursor addressing and
control, a 14 key numeric key pad, 16 pro-
grammable function keys, blinking, a self-
test mode, protected fields and much, much
more.
The new INTERTUBE video terminal from
Intertec represents the latest advances in
microprocessor technology. Speed, silence
and flexibility, coupled with the many oper-
ator-oriented features of the INTERTUBE
serve to boost the efficiency of both soft-
ware and programmers.
Our combination of price and performance
gives a true advantage to a broad cross-sec-
tion of users in the computer industry.
Whether your requirement is the most so-
phisticated data entry application or a sim-
ple inquiry/response environment, the
INTERTUBE is the best value for your ter-
minal dollar.
The INTERTUBE offers a variety of inter-
faces including RS— 232C and 20MA cur-
rent loop. Better yet, the INTERTUBE
interfaces beautifully with the user. Pains-
taking human engineering assures you of
many hours of enjoyable operation without
operator fatigue. INTERTUBE's high res-
olution CRT monitor provides the sharpest
possible display image. And our special
Accu-Dot™ focusing technique produces
ultra clear characters on a non-glare screen.
Our INTERTUBE not only offers unparal-
leled price and performance but is also
backed by Intertec's nationwide factory
trained service network providing over 250
local dealer and service center outlets.
There are lots of video terminals on the mar-
ket today but only the INTERTUBE de-
livers significantly better performance for
such an incredibly low price. So before you
buy one of those dumb terminals or one of
those other smarter ones, you owe it to
yourself to sit down in front of an
INTERTUBE. While we admit our motives
for recommending that you try the
INTERTUBE are purely selfish, it's the only
true way to appreciate the design features
and price advantage of our new terminal.
Mere words just can't do justice to the
INTERTUBE. Contact your local dealer to-
day and discover just how smart you'll be
when you spend only $784 for your new
INTERTUBE.
For the name and address of your nearest
dealer, contact us at one of the numbers
below. It's the only smart thing to do.
*$784 is INTERTUBE's "Quantity One-
Immediate Delivery" price. Dealer in-
quiries are invited.
INTERTEC DATA SYSTEMS
Corporate Headquarters
1851 Interstate 85 South
Charlotte, North Carolina 28208
704/377-0300
Eastern Regional Marketing
1 9530 Club House Road
Gaithersburg, Maryland 20760
301/948-2400
Western Regional Marketing
17952 Sky ParkBlvd.
Irvine. California 92714
714/957-0300
TERMINAL SPECIFICATIONS
MEMORY SIZE:
24 lines by 80 characters per line.
ALPHANUMERIC CHARACTER SET:
Generates all 128 upper and lower case ASCII characters.
LINE DRAWING CHARACTER SET:
Eleven special graphics symbols.
DISPLAY PRESENTATION:
Dark characters on a light background. (Display is reversable through
keyboard selection.)
VISUAL ATTRIBUTES:
Blinking (2 frames per second), underline, reverse video, half intensity.
SPECIAL ATTRIBUTES:
Protected, constant and print-only fields.
STATUS INDICATOR LINE:
25th display line reserved for terminal status messages. (Status line displayed as
inverse of normal display.)
SCREEN SIZE:
12 inch diagonal.
REFRESH RATE:
60 frames per second (50 frames per second— Export model).
OPERATING MODE:
Conversational: character at a time transmission.
Message: line at a time transmission.
Page: full or partial screen at a time transmission.
TRANSMISSION MODE:
Half or full duplex, keyboard and switch selectable.
KEYBOARD:
Standard teletypewriter-compatible layout plus 14 key numeric pad, local
mode and erase. Keyboard lock/unlock under program control. Also, 16
programmable function keys.
SELF— TEST MODE:
Self-diagnostic firmware testing routine-results displayed on status line.
ADDRESSABLE CURSOR:
Direct positioning by either discrete or absolute addressing.
CURSOR CONTROL:
Up, down, forward, backward and home-keyboard selectable.
READ TERMINAL STATUS:
Allows CPU to interrogate for terminal status, present cursor address and
memory value at cursor position.
PARITY:
Choice of even, odd, marking or spacing— keyboard and switch selectable.
WEIGHT:
Approximately 37 pounds.
PHYSICAL DIMENSIONS:
14 5/8" (H) x 21 3/8" (W)x23 1/8" (D).
ENVIRONMENT:
Operating temperature between 0° to 50° C, Storage 0° to 85° C, 10 to 95%
relative humidity, non-condensing.
COMMUNICATIONS INTERFACE:
El A RS— 232C/CCITT V.24 operates through the range of 75 to 19.2K BPS,
switch and keyboard selectable. Available 20/60 MA current loop operates
through the range of 75 to 9600 BPS.
AUXILIARY INTERFACE:
RS— 232C/20MA current loop printer port.
The most remarkable specification of the INTER-
TUBE is its astonishingly low price. The feature-
packed INTERTUBE compares with terminals costing
more than three times as much! Combine this price
with the inherent reliability of our microprocessor de-
sign and the result is an industry first!
The INTERTUBE and other Intertec products are distributed in these cities: Albany, Alberquerque, Atlanta, Baltimore, Boston,
Buffalo,Charlotte, Chicago, Cincinnati, Cleveland, Columbia, Columbus, Dallas, Dayton, Denver, Detroit, Hartford, Huntsville, Houston,
Indianapolis, Jacksonville, Kansas City, Los Angeles, Louisville, Madison, Memphis, Miami, Milwaukee, Minneapolis, Nashville, New Haven,
New Orleans, New York, Oklahoma City, Omaha, Philadelphia, Phoenix, Pittsburgh, Portland, Providence, Raleigh, Richmond, Rochester,
Sacramento, Salt Lake City, San Antonio, San Diego, San Francisco, Seattle, St. Louis, Syracuse, Tampa, Tulsa, Virginia Beach,
Washington, DC. International offices located in Australia, Austria, Canada, Denmark, England, France, Germany, Holland, Switzerland.
$784
* QUANTITY ONE!
You don't have to buy a hundred terminals or sign a
contract to qualify for INTERTUBE's low $784 price
tag . . . it's our "quantity one— immediate delivery"
price. Additional discounts are available under OEM,
GSA/Educational and Distributor pricing schedules.
Contact an Intertec Marketing office for details.
Kenneth S. Widelitz
Attorney-at-Law
Personal Privacy
A discussion of personal pri-
vacy and computers looks at only
that portion of personal privacy
relating to records and record-
keeping practices. This comprises
a very small portion of what is or-
dinarily referred to as personal
privacy.
The Constitution does not
mention a right to privacy.
Nevertheless, various aspects of
personal privacy have been pro-
tected against government action
by traditional interpretation of
the Bill of Rights. The First
Amendment guarantees the rights
of free speech, freedom of assem-
bly and freedom of religion; the
Third Amendment prohibits
quartering soldiers in your home;
the Fourth Amendment prohibits
unreasonable searches and sei-
zures; and the Fifth Amendment
protects against self-incrim-
ination.
Courts have interpreted these
protections to prevent an individ-
ual from being forced to reveal
political, social or philosophical
beliefs. Furthermore, the Su-
preme Court has indicated that
individuals have a right to privacy
relating to their freedom to prac-
tice contraception, have an abor-
tion or read pornography at
home. The judicial system has
also developed principles that
allow suits for invasion of privacy
in certain situations involving
financial or reputational injury
of one person by another.
The foregoing aspects of pri-
vacy have nothing to do with the
notion of invasion of personal
privacy by computer. “Invasion
of personal privacy by comput-
er,” of course, does not mean the
computer is responsible for inva-
sion of an individual’s personal
privacy. It would be more accu-
rate to say “invasion of personal
privacy by use of a computer.”
It is difficult to define personal
privacy in terms of potential inva-
sions by use of a computer. Be-
fore we go on, let’s speculate on
some repugnant uses of a com-
puter to invade what we com-
monly consider our domain of
personal privacy.
You are being interviewed for a
job. After the usual meetings
with key executives, and the
three-martini lunch, a clerk in the
personnel department makes a
critical phone call to Interna-
tional Data Base, Inc., which re-
ports that although you have not
had any prior criminal convic-
tions or arrests, you have ordered
$500 worth of merchandise in the
last year from a firm specializing
in pornographic films.
The report shows that you
bought an additional $200 worth
of materials from an adult book-
store, and made a $1000 contri-
bution to the Communist Anti-
defamation League. It reveals
that you travel without your wife
on business trips, but always
register at hotels as Mr. and Mrs.
It also notes that during your
sophomore year in college you
failed a computer-programming
course. During the last five years,
the report discloses, you have
paid $10,000 to a psychiatrist.
You are surprised when you do
not get the job.
To my knowledge, such dos-
siers are not currently available to
potential employers. However,
virtually all the information from
which such a dossier could be pre-
pared is presently in a computer
data base somewhere.
International Data Base, Inc.,
does not yet exist, but many
credit-reporting agencies do. Fi-
nancial institutions rely on such
agencies to provide information.
It is commonplace to read of per-
sons who have been denied fredit
because of inaccurate reports
issued by credit bureaus. The
cause of such inaccurate reports
is uniformly attributed to “com-
puter error.”
Back to our definition of per-
sonal privacy. From the above, it
can be seen that the invasion of
personal privacy by use of a com-
puter relates to the compilation
and dissemination of informa-
tion. Some information about
you is public, available to anyone
to research and use (e.g., records
of your birth, marriage, criminal
convictions, ownership of real
property, address, etc.). Count-
less other records (educational
transcripts, military records,
employment personnel files,
etc.), though not “public,” are
accessible by numerous other
people.
Even if you shudder at the
thought of a firm such as the fic-
tional International Data Base,
Inc., consider that every piece of
information in their hypothetical
dossier on you is already known
to many other persons (the por-
nography company, the Commu-
nist Anti-defamation League, the
airline-reservation clerk, the
hotel desk clerk). You did not
think in terms of invasion of per-
sonal privacy when you placed
the pom order, gave the contribu-
tion, made the airline reservation
or checked into the hotel.
You can see that in defining the
concept of personal privacy with
relation to records and record
keeping, it is not so much a con-
cern that others know informa-
tion about you, but rather to
what extent that information is
compiled and communicated to
persons other than those who
garnered the data on a first-hand
basis— that is, those persons nec-
essary to consummate a transac-
tion. In this context, the right to
privacy involves the right of an
individual to decide for himself
when and on what terms his acts
should be revealed to the general
public.
As stated by Charles Fried in
an article entitled “Privacy”
(Yale Law Journal , 1968, p. 482),
“It is not true, for instance, that
the less that is known about us the
more privacy we have. Privacy is
not simply an absence of infor-
mation about us in the minds of
others; rather it is the control we
have over information about
ourselves.”
Of course, for any individual,
privacy as a value is not absolute.
Its importance can vary with sub-
ject matter, time, age, etc. It must
also be recognized that the right
of personal privacy can conflict
with the interests of society (i.e.,
records maintained for criminal
investigations or national-securi-
ty problems). A conflict also ex-
ists in the area of free speech and
the public’s “right to know.”
The Problem Historically
Having ruminated about the
definition of personal privacy
and the right to it, vis-a-vis com-
puters, it might be useful to con-
sider from where our notions of
the right to control information
about ourselves come.
From the point of view of the
individual, the right to personal
privacy in terms of control of in-
formation about ourselves is a
relatively recent phenomenon.
Man has kept records of births,
marriages, deaths, etc., almost
since he learned to write. They
were kept on a very local basis be-
cause no methods were available
for communicating their content.
Virtually no other information
about an individual was record-
ed. It wasn’t necessary to ask a
credit-reporting agency if a per-
son was credit-worthy. Everyone
in his village knew that he was or
wasn’t.
Perhaps it was with the inven-
tion of the Gutenberg printing
press, which made the wide-
spread dissemination of informa-
tion possible, that the question of
the existence of a right to person-
al privacy was first raised.
Subsequent technological ad-
vances resulted in the rapid com-
munication of information over
long distances. Telegraph, tele-
phone and radio made the world
smaller and information on per-
sons outside our communities
more important. That technology
gave us the means to invade the
personal privacy of a limited
number of individuals, who tend-
ed to be the “important” people,
public people, such as politicians,
actors and actresses and com-
munity leaders. Although it was
possible prior to the computer
age to accumulate information
on individual nonpublic persons,
the amount of effort involved in
terms of compiling such informa-
tion was relatively large.
With the advent of the com-
puter, the task of compiling enor-
mous amounts of information on
an enormous number of people
became possible and economical-
ly feasible. Once the capability
existed, the applications followed
closely behind. It is with the ap-
plications, such as credit report-
ing, that the issue of the right to
personal privacy and its invasion
comes to light.
Historically, the social utility
of record-keeping practices has
seldom been questioned. Certain-
ly census information and other
statistical records, along with
some intelligence (i.e., police,
CIA) and administrative records,
are valuable and useful tools that
have a positive effect on society.
Privacy and Personal Computers
The foregoing discussion has
(continued on page 21)
12
KB CLUB
CALENDAR
Steve Fuller
Richardson/Dallas/Ft. Worth TX
Here are the highlights of a let-
ter I received recently from Neil
Ferguson, president of The Com-
puter Hobbyists Group of North
Texas:
One of the oldest computer
hobbyist clubs in the country,
CHG-NT holds two meetings per
month — one for the Richardson/
North Dallas area and another
for members in Dallas/Fort
Worth. Meetings are conducted
on the first and third Saturdays of
the month at 1 pm unless the dates
conflict with holidays. Dues of $7
per year include a subscription to
the club newsletter, “The Printed
Circuit.”
The June meeting will be the
club’s annual swap meet, known
as the “Chip and Dip Fest” in
honor of the IC. Sounds great!
For more information on the
club and its activities, write R.
Neil Ferguson, PO Box 1344,
Grand Prairie TX 75051, or call
him at (817) 265-9054.
Atlanta GA
George Reeves asks us to an-
nounce that the Atlanta Chapter
of the SOL Users Society
(SOLUS) meets twice monthly,
on the first Monday and third
Thursday. Consider it done,
George.
Incidentally, time and space
don’t permit publication of club
rosters, but thanks for the copy.
I’ll know who to call next time
I’m in Atlanta!
Send your requests for details
of the current month’s club ac-
tivities to George at 5002 Crowe
Dr., Smyrna GA 30080, or call
him at (404) 436-0718.
Pensacola FL
The second Thursday of each
month is the meeting date for the
North Florida Computer Society,
according to Eugene Rhodes.
Write to him at 227 Edison Dr.,
Pensacola FL 32505, or call (904)
453-3844.
Ojai CA
W. P. Dart of 231 Valle Rio
Ave., Ojai CA 93023, wants to
hear from hobbyists in his area
who are interested in starting a
computer club. His phone
number is (805) 646-5824.
Augusta GA
The CSRA Computer Club
meets on the third Thursday of
each month at 7 pm in the main
auditorium of the Augusta-Rich-
mond County Public Library in
Augusta.
Rolston Wilder, 2704 Rose-
wood Court, Augusta GA 30909,
will send you a copy of the club
newsletter. For more informa-
tion, write to Rolston or call (404)
733-8750.
Rochester NY
KIM-1 owners in the Rochester
area are invited to join Murray
Smith in forming a club. He says
he suspects there are at least 20 of
you out there, so get in touch with
him at 1972 E. Main St., Roches-
ter NY 14609.
Investors Club
The Microcomputer Investors
Association, a nonprofit profes-
sional group, has been formed to
facilitate the exchange of infor-
mation relating to microcomput-
ers and investments including
stocks, bonds, stock options and
commodities.
In order to benefit from the ex-
periences of others there is a basic
requirement that, at least once
each year, each member submit
an original article for publication
in the association’s newsletter.
If you’d like an application,
send an SASE to Jack Williams,
The Microcomputer Investors
Association, 2415 Ansdel Court,
Reston VA 22091.
Toronto, Canada
Congratulations to the Toron-
to Region Association of Com-
puter Enthusiasts (TRACE),
which celebrated its second birth-
day in February.
According to public-relations
secretary Ross Cooling, “TRACE
began in February 1976 as an in-
formal meeting of ten people in-
terested in personal computing,
and quickly attracted followers.
Since that time the club has
grown to approximately 100
members and about as many ca-
sual followers. Nearly half of the
members have personal systems
of some form.
‘ ‘The main purpose of TRACE
is to foster communication and
resource sharing among com-
puter hobbyists and profes-
sionals. The meeting format in-
cludes one or two talks on micro-
computer-related topics, and
usually a system demonstration.
The club also has a monthly
newsletter, group purchasing and
a library of product literature,
books and periodicals. Other ac-
tivities include flea markets, ex-
hibitions and a software library.”
The club meets at the north
campus of Humber College at 8
pm on the fourth Friday of the
month and at the Ontario Science
Centre on the second Sunday of
the month. Newcomers are wel-
come, and information may be
obtained from Ross at 488-3314,
or Gifford Toole at 828-9202.
Houston TX
Information concerning activi-
ties of the Houston Amateur
Microcomputer Club (HAuCC)
is available from Clifford Carley,
corresponding secretary, PO Box
37102, Houston TX 77036; you
may call him at 921-7532.
The club’s newsletter, “Nyb-
ble,” contains schedules of up-
coming events, product briefs
and computer-related job open-
ings in the Houston area.
Columbus OH
The Amateur Computer Soci-
ety of Columbus (ACSC) meets
at 7:30 pm on the first Wednesday
of each month at the Center of
Science and Industry.
President Fred Hatfield notes
that ACSC is working with the
Columbus Public Library to set
up a “Computer Corner” where
patrons will have access to two
Commodore PET computers and
an extensive technical library.
According to a brochure from
the public library, through “the
generosity of the membership of
the Columbus Amateur Comput-
er Society, the Public Library of
Columbus is now able to offer re-
source materials to the growing
number of computer hobbyists in
the Franklin County area. Mate-
rials ... are available for loan at
the Business and Technology
Division of Main Library, 96 S.
Grant Ave.”
You can contact the Society
c/o Fred Hatfield, Computer
Data Systems, 1372 Grandview
Ave., Columbus OH 43212, or
call (614) 488-3347.
Granger IN
A software/hardware library
to support the Apple-I computer
has been started here. Interested
persons may write to Joe Torzew-
ski, 51625 Chestnut Rd., Granger
IN 46530.
Midland/Odessa TX
The Permian Basin Computer
Group has scheduled meetings of
its two chapters as follows:
The Midland meetings will be
held on the second Monday of
each month at 7:30 pm in the Stu-
dent Center on the Midland Col-
lege campus. Meeting dates are:
May 8, June 12, July 10, August
14, September 11, October 9,
November 13 and December 11.
The Odessa chapter will meet
on the second Saturday of each
month at 1 pm in Room 203 of the
Electronics Technology Building
on the Odessa College campus.
Dates are: May 13, June 10, July
8, August 12, September 9, Oc-
tober 14, November 11 and
December 9.
Details are available from the
Permian Basin Computer Group,
c/o John Rabenaldt, Ector
County School District, Box
3912, Odessa TX 79760. The
phone number in Midland is (915)
697-4607 after 6 pm; in Odessa,
(915) 332-9151 Ext. 43 from 8 am
to 5 PM.
Cornwall, England
British hobbyists may obtain
information about the Personal
Computing Club (PCC) from
(continued on page 21)
13
„ NEW
Products
Two New Mainframes from TEI
TEI, Inc., has two versions of
their Computer Mainframe Sys-
tem available. The first is Model
MCS-112, a foundation unit
based on an S-100 bus system
with a 12-slot motherboard. The
power supply is rated at 17 Amps
at 8 V and 2 Amps at ± 16 V.
The second version is Model
MCS-122, a foundation unit also
featuring the S-100 bus system
with a 22-slot motherboard and a
higher rated power supply. Both
models are housed in a heavy-
duty, vented aluminum cabinet.
The power supplies consist of a
constant voltage transformer
providing brownout protection
and showing high noise immunity
of better than 100 db between in-
put and output. The front panels
of both fully tested and assem-
bled models include an indicating
ac switch and reset switch. Prices
are $395 for the MCS-112; $495
for the MCS-122.
For more information, contact
CMC Marketing Corp., 5601
Bintliff, Suite 515, Houston TX
77036.
INFO 2000 Systems Software
Supported for 8080 and Z-80
INFO 2000 disk-systems own-
ers may now utilize expanded
software for both 8080 and Z-80
microcomputers. INFO 2000
Corp. has increased disk-operat-
ing software capability using Dig-
ital Research CP/M. Among the
software packages that operate
under CP/M are two full versions
of Disk BASIC, including Micro-
soft Extended Disk BASIC
(4.41), $350. Also available is a
Structured Systems Group
QSORT ($95), a high-speed, gen-
eral-purpose sort package. A
name and address maintenance
system, NAD ($79), includes
mailing labels. The manufacturer
states that all software is current-
ly available for immediate de-
livery for use on INFO 2000 Disk
Systems.
Still available for Z-80 systems
is the complete TDL software
package— including 12K BASIC,
macro assembler, Z-TEL Text
Editor and Word Processor— ex-
panded to operate under CP/M
and Zapple operating systems.
Price, $215.
All INFO 2000 software oper-
ates on INFO 2000 Disk Systems
that employ the PerSci 277 dual
diskette drive with intelligent
controller. The Disk System is
available for all S-100 microcom-
puters using Z-80 or 8080 proces-
sors, for Digital Group Z-80 and
8080 systems, and for the Heath-
kit H8.
INFO 2000 Corp., 20630 S.
Leapwood Ave., Carson CA
90746.
New EPROM Programmer
Smoke Signal Broadcasting,
PO Box 2017, Hollywood CA
90028, announces a new, low-
cost 2708 EPROM programmer.
Designated the POP-1, the unit
lists for $149 and is designed to
interface to the company’s P-38-1
and P-38-FF EPROM boards,
which are SS-50-bus-compatible
products. Complete software is
provided on audio cassette. A
unique adaptive programming
technique allows most 2708s to be
programmed in 15 seconds in-
stead of the usual 1 Vi minutes. A
separate, self-contained power
supply is used for the program-
ming voltage to insure sufficient
current capability to program
EPROMs from any manufacturer.
Microprinter with
Serial Interface
Centronics Data Computer
Corp. introduces the Microprint-
er-Sl with serial interface. The
new printer, which has a single
unit price of $695, allows the user
to select baud rates, parity and
the number of stop bits.
Centronics expects the SI ver-
sion of the Microprinter to enjoy
an even broader range of applica-
tions than the PI , which debuted
last June. Both units are aimed at
the home, hobby and micropro-
cessor markets, and both are
suited for use in diagnostic sys-
tems, CRT hard-copy applica-
tions, industrial instrumentation,
and demand message printing.
However, since many CRT ter-
minals require the RS-232 inter-
face, the SI should see frequent
use as a remotely placed message
printer.
Information about numerous
other features of the SI (and PI)
can be obtained from Centronics
Data Computer Corp., Hudson
NH 03051.
CP/M on North Star Disk
CP/M has become the most
widely used S-100 floppy-disk
operating system, effectively
making it the software-exchange
bus for S-100 systems. Vendors
of hardware or software support-
ing CP/M include Imsai, Cro-
memco, TDL, Digital Systems,
Tarbell Electronics, Microsoft,
Digital Research and many
others. Also, the CP/M Users’
Group is currently operating a
successful software exchange
program.
Now North Star Disk users can
also join the software bus. With-
out any hardware changes, CP/M
can be run with all the features
available to the users of the
system on standard floppy disks.
Microsoft FORTRAN-80 and
Disk Extended BASIC can also
be supplied on a 5 14 inch diskette
to run on “CP/M on North Star
Disk.’’ All the software is fully
8080/Z-80 compatible.
Retail prices: CP/M on North
Star Disk, $112; FORTRAN-80,
$400; Disk Extended BASIC,
$300. FORTRAN-80 includes re-
locating assembler and linking
loader. Full ANSI except for
complex data types. Disk Extend-
ed BASIC is a CP/M generation
of Altair Disk BASIC 4.1. All
prices are freight inclusive.
Lifeboat Associates, 36 W.
84th Street, New York NY 10024.
14
_3f V_ GRAPHICS
' ^ RRE
Typefaces may even be MIXED on the
same line.
The Axiom 801 and sample printout from the 820 ($795).
The Axiom Printer
Axiom has a printer, the EX-
801 , part of the EX-800 series, for
fast and inexpensive printing. It
prints on aluminized paper ($4
per 230-foot roll) by passing a
current through the aluminum
coating, evaporating it and thus
leaving the black-coated paper
underneath exposed. Moisture
and sunlight have no effect on the
paper. The current is passed
through eight wires, making a
dot-matrix pattern without the
need for moving wires and their
fragility.
The printer, with an 8048
microprocessor in control, has an
input buffer of 256 characters
which can be expanded to 2K,
thus providing a full CRT page
dump at one time. It will work
from RS-232, 20 mA serial as well
as parallel ASCII. It comes with
96-character ASCII standard and
is expandable to 256 characters
with your own programmable
fonts. Three character widths,
which can be software-selected to
provide 20, 40 or 80 columns on
the 5-inch-wide paper, can be
mixed on a line for emphasis.
Axiom Corp., 5932 San Fer-
nando Rd., Glendale CA 91202.
Audio Cassette Magazine
CLOAD Magazine, published
by R D McElroy & Co. of Goleta
CA, and written especially for
Radio Shack’s TRS-80 computer,
is something new. It is “printed”
on a standard audio cassette and
will load directly into the TRS-80
computer. Its “articles” are pro-
grams ranging from short games
to involved programs of a practi-
cal nature; emphasis will be on
education.
People can submit programs for
publication. There will be 12 is-
sues a year, each issue consisting
of an audio cassette with six to
ten programs (more, if possible),
an index and an instruction sheet.
Subscription rates available.
CLOAD Magazine, Box 1267,
Goleta CA 93017.
New Business Software Program
A new, low-cost ($200) gener-
al-business software package for
microcomputers that includes
general ledger, accounts receiv-
able, accounts payable, finished-
goods inventory control and pay-
roll has been introduced by The
Computer Mart, Orange CA.
Developed by Larry G. Grimes
& Associates and Computer
Products of America, a subsidi-
ary of The Computer Mart, the
24K Grimes Business Informa-
tion System (GB1S) is specifically
designed for small businesses. In
a typical application, the GBIS
can store on a single minifloppy
diskette up to 400 customer list-
ings, 50 vendors, 400 line items of
inventory, 25 employee records
and 60 general-ledger accounts.
GBIS is written in North Star
BASIC, although other disk
BASIC languages can be used for
the listings. Programs are written
so that someone with a minimal
understanding of the BASIC pro-
gramming language can use it.
The Computer Mart, 633 West
Katella Ave., Orange CA 92667.
4 MHz Z-80A Processor Board
The ZPB 4 MHz Z-80A proces-
sor board is one of several new
high-performance products
North Star offers for S-100 bus
computers. The ZPB will operate
in systems with or without front
panels. It includes space for IK of
2708 EPROM.
Other features of this versatile,
fast board include auto-jump
start-up and vectored interrupts.
Available as a kit for $ 1 99 or $259
fully assembled. EPROM option
costs are $49 for kit and $69
assembled.
North Star Computers, Inc.,
2547 9th Street, Berkeley CA
94710.
Video Checkers on Cassette
Compu-Quote, 6914 Berquist
Ave., Canoga Park CA 91307,
has developed several ?ames on
cassettes, recorded in the Tarbell
format and programmed in Mits
BASIC (3.1). Contained on one
cassette, Video Checkers pro-
duces checkerboard graphics on
the CRT when used with the
PolyMorphic Video Interface
and 64-character option. Two
versions of the program on one
60-minute cassette play a chal-
lenging game that conforms to
International Rules. The first ver-
sion requires a total of 16K of
memory, inclusive of 8K BASIC.
The second version is more
graphic and requires an addi-
tional 4K.
As the player and computer
each take turns, the checkers
blink and move to indicate their
(continued on page 114)
15
151 BOOKS
An Introduction
to Microcomputers
Volume 0, The Beginner's Rook
Adam Osborne
Adam Osborne and
Associates, Inc.
PO Box 2036
Berkeley CA 94702, 202 pages
Hallelujah! At long last, some-
one has actually produced a book
that really is for the absolute
novice in computer technology.
Like many books before it, this
one claims to be written for peo-
ple who know nothing about
computers. The novelty is that
Osborne et al mean it; and what is
even better, the book is for the
potential user who doesn’t know
how a computer works, will prob-
ably never know and doesn’t par-
ticularly care. It is written for a
pure user orientation, treating the
microcomputer as an appliance
and recognizing that it is possible
to use the appliance without
knowing how to design it, modify
it or repair it. The reader need
only know that the microcom-
puter exists, that it is potentially
capable of performing interesting
functions and that he’d like to use
one.
Starting from those basic
assumptions, the book takes the
reader logically and painlessly
through a survey of the current
state of microcomputers. The
first chapter does nothing more
than identify the major boxes
that make up a microcomputer
system and explain some basic
data recording concepts. Chapter
2 introduces the book’s hero, the
indomitable Joe Bitburger. In re-
counting Joe’s experiences with
his first basic kit, Osborne de-
scribes the frustrations of operat-
ing with only front-panel switch-
es and lights for input/output.
This brings up consideration of
the Teletype and of keyboard en-
try in general. That consideration
leads logically into Chapter 3,
which introduces the concept of
logical units (files and records),
as opposed to physical units, and
discusses various components
and devices in considerable depth.
Chapter 4 is the obligatory dis-
cussion of binary/octal/hex
arithmetic and logical opera-
tions. There are numerous ex-
amples, and the explanations are
unusually lucid.
Chapters 5 and 6 deal with soft-
ware and with system organiza-
tion. The software discussion de-
fines those mysterious terms— as-
sembler, compiler, interpreter —
clearly and concisely. With those
concepts established, the chap-
ters describe the various ap-
proaches to machine organiza-
tion. The advantages of different
word lengths, of different num-
bers and arrangements of regis-
ters, and other computer ele-
ments are clearly explained. The
author’s intent in these chapters
is to bring the reader along to a
point where he can intelligently
read the specifications of various
devices and systems and evaluate
these in the light of his own par-
ticular requirements.
At least one other aspect of The
Beginner's Book deserves not
only mention, but unrestrained
praise: It is readable, even enter-
taining. The eager, but frequently
misguided, Joe Bitburger is an
engaging hero. He helps to keep
the text oriented to a user (rather
than a designer), and to a rank
beginner at that. The generous
use of graphics and cartoons aids
in understanding the text and
demonstrates (if it was required)
that visual aids and humor can in-
crease the impact of even tech-
nical writing.
Less this sound like a commer-
cial for Osborne and Associates, I
should mention that the book
does have a few flaws. As an ex-
ample, on pages 5-21, the text
refers to a drawing and says that
the number representing an in-
struction code is circled. In fact,
there are no circles on the draw-
ing and the referenced number is
in a shaded block. This sort of
thing can be confusing, but there
are few such errors.
A second objection is that the
very readable style of the early
chapters is not carried all the way
through the book. By the end of
Chapter 6, Mr. Osborne has ba-
sically reverted to the technically
thorough but very dry style of his
other books. In his defense: He is
dealing with more complex data
at this point. In truth, there
would be no objection to this if he
hadn’t done such a damn fine job
in the first part of the book.
In summary, Mr. Osborne has
delivered a first-rate book for the
novice computer hobbyist. It is
successfully aimed at that great
throng of people who would like
to use computers but care no
more about the design and repair
of such devices than they do
about the internal workings of
their refrigerators. This is a su-
perb book for any beginner and
would serve very well as the text
for an introductory class at any
level from junior high school on
up.
Art McDonough
El Segundo CA
Practical Microcomputer
Programming: The Intel 8080
Weller, Shatzel, Nice
Northern Technology Books
Box 62, Evanston IL
60204 $21.95
Want to know how your 8080
microcomputer really works?
Thinking of writing or modifying
system software? Tired of
BASIC? Ever wondered about
the “Playboy” effect in decimal
arithmetic? A necessary step is to
learn assembly language, and this
book is for you.
Throughout the 18 chapters
and 306 pages, the reader is intro-
duced to hardware features of the
8080, presented in a clear, con-
cise, logical manner. Over 80 ex-
amples are included, most of
which are usable as subroutines
to programs you will want to
write later. Complete chapters
are devoted to such topics as
stack pointers, arrays and tables,
binary/decimal conversion, com-
munications with terminals, in-
terrupts and subroutines. Sixteen
pages are devoted to debugging!
A powerful plus for this book
is the authors’ determination to
demonstrate why and how to use
each instruction, not merely to
explain how it works. Examples
are described first in a short para-
graph, then presented in program
format with instructions printed
in bold letters for easy readabil-
ity. No matrix-printer-with-a-
weak-ribbon printing here!
Reader interest is maintained
through occasional use of humor;
interchanging memory locations
named MOUNTN and
MUHAMD or adding the con-
tents of two double-precision
fields named INSULT and IN-
JURY help make the learning
fun. In a chapter on controlling a
complex peripheral (the Victor
matrix printer), examples include
routines to transform ASCII to
required bit patterns for printing
the Cyrillic alphabet. This new
code (named by the authors) Rus-
sian Unified Standard Key for In-
formation Interchange (RUSKII)
is then printed. Understanding
this chapter will leave the hard-
ware freaks with little fear of
tackling an APL printer.
At no point do the authors
resort to rehashing material avail-
able from the manufacturer of
the 8080, but instead choose to
take a less theoretical, more prac-
tical approach. Oh yes, there is
even an index to allow use as a
reference manual. One minor ob-
jection is that the examples do not
strictly follow manufacturers
standards for assembly-language
programming — labels are not fol-
lowed by a colon and comment
lines begin with a rather than
a The style is similar to
assembly language used on larger
machines.
Any 8080 or Z-80 user would
do well to purchase this work, for
learning and for reference.
Tim Turner
Spokane WA
Take a Chance
with Your Calculator
Probability Problems for
Programmable Calculators
Lennart R#de, Dilithium Press
PO Box 92,
Forest Grove OR 97116
Take a Chance with Your Cal-
culator will help you understand
basic probability theory by run-
ning experiments on your pro-
grammable calculator.
The book is divided into three
sections. The first is composed of
eleven chapters containing a total
of 143 problems to be investigat-
ed, along with some explanations
of the problems and the methods
to be used in solving them. The
second section has 46 pages of
more detailed write-ups on many
of the problems. Section three
has over 100 programs, written
for Hewlett-Packard HP-25-type
calculators, which are used in
solving the problems from sec-
tion one.
Section one contains the real
meat of the book. It starts with a
method of generating random
numbers and then gives methods
16
of testing random-number gener-
ators. These are followed by
chapters dealing with the genera-
tion of specific types of random
numbers, such as the number of
spots showing when you throw a
pair of dice.
Chapter 5 investigates assorted
problems dealing with random
numbers, and other problems in-
volved in drawing marbles from
an urn containing both black and
white marbles. Chapters 6 and 7
deal with towers made of blocks
and the results obtained when
you move a random number of
blocks from one tower to another
randomly selected tower.
Chapter 8 examines “runs”
that are sequences of the same
random digit repeated several
times in a row. Chapter 9 is about
random walks, when you ran-
domly walk either right or left a
random number of paces, and
other extensions to this type of
walk. Chapter 10 defines statis-
tics terms, including consistency
of an estimate, efficiency of an
estimate and confidence level,
and exhibits test problems.
Finally, Chapter 11 comprises
more miscellaneous problems in-
cluding family planning, the
birthday problem, simulation of
an election and queuing problems.
Section two consists of com-
ments on the problems in section
one. Not all problems have com-
ments, and some comments are
quite brief while others are
lengthy. The comments may in-
clude an explanation of the prob-
lem, mathematical formulae, dia-
grams, flowcharts or hints on
how to program a solution for the
problem. In many cases, a refer-
ence is given to another book
where a more detailed treatment
of the problem may be found.
Section three has 79 pages of
calculator programs and instruc-
tions relating to their operation.
The programs were written to
work on HP-25-type calculators.
The book claims they will also
run on Texas Instruments SR-56
calculators; but I believe some
modification to the programs will
be needed since the programs ap-
pear to be written for Reverse
Polish Notation machines. Texas
Instruments uses algebraic entry.
The overall organization of the
book is rather poor. I constantly
jumped from the problem state-
ment in section one to the com-
ments in section two to the pro-
gram in section three, then back
to section one again. If all of the
material for a given problem were
collected in one place the book
would be much easier to use.
I think the greatest fault with
the book is that the author didn’t
seem quite sure just how ad-
vanced his reader would be. In
some cases explanations are given
in detail, while in others there is
no explanation at all. For exam-
ple, problem 23 asks you to com-
pute the standard deviation. Sec-
tion two gives a formula for this
calculation; but having calcu-
lated the standard deviation,
nowhere are you told what it rep-
resents or what it can be used for!
Even with its faults, there is
still a lot of useful material in
Take a Chance with Your Calcu-
lator. The beginner will certainly
learn some new concepts; the
reader with more background
may be given a more intuitive feel
for the theory after observing it in
practice. On the other hand, if
you are trying to learn probability
from scratch, you will need a bet-
ter textbook than this.
Glen Charnock
Oxnard CA
Computers, Computers,
Computers: In Fiction
and in Verse
D. Van Tassel, editor
Thomas Nelson Inc.
30 E. 42 St., NYC
$6.95, 192 pages
Dennie Van Tassel, author of
several successful books for com-
puting professionals, has been
gradually moving into broader
areas. The Compleat Computer
(reviewed in Kilobaud No. 1) was
a tasty assemblage of semitechni-
cal articles, news items, cartoons
and some fiction, organized
around specific issues of interest
to those attempting to under-
stand the effects of computing on
our society. Now, in Computers, I
Computers, Computers, he has
pulled together a wide assortment
of Fictional treatments of com-
puters. The result can be enjoyed
equally by both computerniks
and nontechniks.
The book’s 18 short pieces of
fiction fall into three categories.
The majority (1 1) are science-fic-
tion short stories by such authors
as Robert Sheckley, Gordon
Dickson, Robert Heinlein and
Barry Malzberg. (In my opinion,
Malzberg’s contribution, “The
Union Forever,” is his most suc-
cessful short piece.)
Another category is satire, and
here we find pieces by such lumi-
naries as Art Hoppe, Art Buch-
wald and Bob and Ray. The third
category, poetry, includes 18
computer-tinged limericks
(“Glorobots” by Gloria Max-
son), as well as several more ex-
tended poems and a song (“Push
the Magic Button” — to the tune
of “Puff the Magic Dragon”).
Although the stories, articles
and poems can be read and ap-
preciated separately, it seems
more interesting to view the col-
lection as a whole. In so doing,
try to extract from the different
treatments some common notion
of what nontechnical people have
thought computers will do to us.
The title of the author’s introduc-
tion says it— “Make Way for the
Machines.”
Rich Didday
Santa Cruz CA
From Don Tarbell . . .
It seems that many new com-
puter customers have been given
the impression that it’s easy to get
a computer system up and going.
The truth is that it can be a can of
worms, depending on the user,
the components and the system
required.
There is one thing in particular
that many people don’t seem to
understand: There is a world of
difference between a bunch of
hardware and software computer
components and a computer sys-
tem up and working and doing
what you want it to do. This dif-
ference is sometimes big, some-
times small, but it’s almost al-
ways at least ten times as big as
you think. This difference is
called system design and integra-
tion, and there are companies
that make millions of dollars do-
ing things like it. This phase is
when you discover that all “S-100
compatible” boards are really
not compatible with each other.
It’s when you find out that two
big expensive pieces of software
can’t talk to each other. It makes
people work weeks or months
without getting anywhere. In
short, you stand to lose your shirt
if you don’t know what you are
doing!
Everyone runs into system in-
tegration sooner or later, but it
seems to get really nasty in small-
business systems. I suppose one
reason is that it all sounds so sim-
ple! The boards are proven, the
software is extensive. Everything
required sits right there in front
of you. All you need to do is
make a few patches here and
there and write a little applica-
tions software.
Of course, it’s those little de-
tails that get you every time. Since
I get many calls from people
about their problems, I’d like to
present short answers to a few
that seem to recur.
1. Many Z-80 CPU boards
don’t create a true Altair-bus en-
vironment. This means that al-
though the CPU board will work
in some situations, with some
memory and with some inter-
faces, it may not work with the
combination you want.
2. Most 8080 software is writ-
ten to run at a particular location
in memory. If two programs that
must run together are written to
run in overlapping memory, one
must be moved. It is not always
easy to move programs, particu-
larly if there is no source available.
3. Most video boards require a
good-sized driver program. This
driver must be present in memory
concurrently with your BASIC
interpreter, or whatever, and fur-
thermore, each program that
must access this drive (probably
all) will have to have references to
this driver in it.
4. All dynamic memory boards
require a refresh cycle. Even
though some manufacturers will
claim theirs is done at a time that
is “invisible” to the rest of the
system, there is no guarantee that
some other manufacturer’s board
won’t need to access memory at
this same “invisible” time.
5. Some programs that work
fine in your main computer mem-
ory (RAM) may not be easily
modified to work with a mass-
storage device such as a floppy
disk or cassette. Often a whole
new approach is required.
6. Noise is almost always a
problem. Some S-100 buses have
more noise than others. Termi-
nating resistors don’t always
help. The third prong on one wall
17
TO 5
(IC8-II)
TVT n
CT-CA
COMPUTER
CONTROLLED
CURSOR
BOARD
©"©"
©-©-
©-®-
©-®-
-Hv
-N/C
- N/C
-N/C *
,0c 2 I5 fl
svn r
. ^FS
- ESC
I
NUL
SOH
STX
ETY ^
EOT -
li
END -
ACK -
ia.
*—• CR
, -FF
HT
»— BS
BEL
plug may not be connected to the
third prong on another wall plug
(would you believe 50 volts of
difference?).
7. Be sure that there are no two
interfaces that use the same port
numbers.
8. There is no real standard on
console port numbers, so either
software or hardware may need
to be modified. Sometimes the
hardware is easier.
9. ASCII is not always ASCII,
RS-232 is not always RS-232, etc.
10. Some mainframes that
don’t have a front panel also
don’t provide some of the re-
quirements of the bus that a front
panel would have, such as pull-up
resistors, MWRITE line, etc.
1 1 . Because every single com-
ponent in your system tests good
in someone else’s system, that
doesn’t mean they will all work
together. The best test is the
whole system the component will
work in.
12. There is no memory test
that goes through all the com-
binations, regardless of what
anyone might say. Figure it out
yourself if you must, but such a
test would take longer than your
lifetime to run. After all, what is
2A(4096*8), anyway . . . and
then, of course, there’s the bit
that drops out after time . . .
13. Programs always take at
least five times longer to write
than you most conservatively es-
timate, even when the hardware
works perfectly all the time,
which, of course, it won’t.
14. Requirements for software
should be specified before those
for hardware. Design the system
from the top down if possible.
The preceding is not meant to
give everyone a crash course in
system design, but only to point
out that there may be problems
you haven’t thought of. If you
are a hobbyist, you are probably
willing to learn (sometimes the
hard way). If you’re a systems de-
signer, be sure to keep on learn-
ing. And if you’re a businessman,
find someone who really knows
what he’s doing to help you out.
It may cost you as much as you
paid for the hardware and soft-
ware (or more); but, believe me,
if you don’t have the time or in-
clination to learn it yourself, the
money spent is well worth it!
In the past several months, I’ve
talked to many people who are in
trouble now because they under-
estimated the task they under-
took. These people would have
been better off if they had re-
tained someone who already
made the mistakes a few years
ago. The computer stores around
the country, especially the older
ones, are the best places to find
people like this. They’ve had the
experience to know which boards
work with which, and which soft-
ware is useful. If you want to
shop around and buy one piece
here and one piece there, directly
from the manufacturers, you’re
taking on the job of designing the
system yourself. So don’t blame
the manufacturers or the stores
where you bought these separate
items for your problems. If
you’re not willing to take the sys-
tem’s responsibility (and that in-
cludes maintenance) on yourself,
find someone who will.
Some manufacturers claim to
be selling computer “systems.”
Watch out! Is the manufacturer
local? Does he provide a good
warranty on the complete sys-
tem? Will the software he pro-
vides really do the job? My feel-
ing is that only the largest manu-
facturers, such as IBM, DEC
and UNIVAC, can afford to pro-
vide local support for complete
business systems of any size (even
the smallest). That’s why I’ve
chosen to avoid the word “sys-
tem” in any of my advertising
and refer as much business as
possible to a customer’s local
dealer. I would rather sell mostly
to true computer hobbyists,
dealers and system designers.
I hope that the above has
helped someone. I don’t want to
discourage anyone from getting a
computer system. I use one in my
own business and feel that it’s one
of the most valuable assets I have.
I just want people to have enough
information to buy at the right
place.
Don Tarbell
Tarbell Electronics
Many thanks , Don. This is a
problem we’ve been trying to
bring out in the open for some
time. Now, if we can just get
some articles into Kilobaud deal-
ing with the problems encoun-
tered in system integration
... the benefits will be tremen-
dous. — John.
Response to
October Article
As a TVT-II owner, I read with
interest Richard Wright’s article
C Kilobaud , October, 1977) de-
scribing ways of decoding the
much needed additional ASCII
control codes. The circuit shown
in Fig. 1 should not have been dis-
missed so quickly, though, as the
substitution of a 74159 open-col-
lector decoder for the 74154 over-
comes the “one flaw” mentioned
by the author. By leaving the
TVT-II Cursor Control board in
place, removing the 7445 and us-
ing the bit signajs^available_at its
socket plus the 4, 5 and 5 sig-
nals available elsewhere on the
board, a 32-character, strobed
decoder can be made using only
two 74159s and half of a 7404.
The enclosed schematic shows the
arrangement I am presently using
in my TVT-II. The page select,
cursor control and outside-world
commands are implemented in
basically the same manner as
shown in Fig. 3 of Mr. Wright’s
article using this simplified
decoder.
Roger Wileman
Escondido CA
Algorithms and
Flowcharts, Please
As a KIM-1 owner and graphic
programmer by trade, I feel jus-
tified in donning my cloak of
righteous indignation when con-
fronted with Larry Woods’ bio-
rhythm program {Kilobaud No.
14, February 1978).
In any program, one must
compromise between the two
mutually exclusive ideals— small
size and quick execution. Al-
though a biorhythm algorithm is
trivial, Larry’s program is not.
The DAY # algorithm in the
algorithms I have provided will
work for any date between 1
March 1900 and 28 February
2100. I found it in an old com-
puter journal. As a side effect,
MOD (DAY # (D, M, Y), 7)
yields the day of the week— SUN
:: = 0, MON :: = 1, etc., so why
all the tables?
I share these algorithms for
two reasons: (1) to try to stem the
flood of biorhythm programs be-
ing published and (2) to demon-
strate the proper way to share
algorithms (i.e., in an implemen-
tation-independent form). Take
note that an algorithm is not a
program (all apples are fruit, not
all fruits are apples).
Though the exchange of pro-
grams should be encouraged, the
6502 owner derives little from a
Z-80 chess program, much from
the algorithm for collecting the
set of squares containing pieces
that can be captured by either
color.
An algorithm can be pro-
grammed in any computer lan-
guage— BASIC, PASCAL, Z-80
or 6502 assembler. An algorithm
is more machine independent
than a BASIC program.
I feel cheated by every pro-
gram that is published in assem-
bly-language for a machine I’m
not familiar with— I must try to
conjure the algorithm from the
code in order to make use of it,
and try not to get sick at a KIM-1
program that can’t be placed in
ROM because it’s self-modifying.
But to get back to my original
purpose for writing — I feel that
the editors must take more re-
sponsibility in ferreting kludges
from this publication. You could
start by requiring that a set of
flowcharts accompany every
assembler program.
Starting with a published al-
gorithm, rather than a machine-
dependent program, each reader
can write a program for his own
machine that takes advantage of
its unique environment (I/O rou-
tines, math software, addressing
modes).
I don’t want to see a Compu-
Craft micro APL program for the
game of checkers with a Cyber-
Clops CRT/TTY MK IV, but I
18
might try my hand at program-
ming go-maku for my KIM-1 if I
had an algorithm in a structured
psuedo-language.
D. A. Harrod
Rochester NY
As a matter of policy we always
try to insure that flowcharts are
included with software articles.
Sometimes we slip up, too. As
far as Larry Woods * article is
concerned, it works. As a matter
of fact, I ran it on the day I re-
ceived your letter and it said I
was at an emotional peak. Your
letter took care of that! — John.
computer user, one that was long
overdue. One wonders why some
of the systems manufacturers
don’t pick up the ball and con-
tribute to the enhancement of
their systems with such “innova-
tions.”
Mr. Parsons’ article, “Under-
standing Loaders,” was an ex-
ceptionally well-written and time-
ly article. His portrayal of a sys-
tem’s software as a series of load-
ers, one bringing in the next in a
hierarchical manner, is a very im-
portant concept for users to
grasp. One can even say (stretch-
ing his analogy to its limits) that
the software of a system is only
BIORHYTHM: PROC;
AGE = DAY # (TDAY, TMON, TYR)-
DAY # (BDAY, BMON, BYR);
PHYS = MOD (AGE, 23)
EMOT = MOD (AGE, 28)
INTL = MOD (AGE, 33)
END BIORHYTHM;
DAY # (D, M, Y)
IF M > Z
THEN DO
K = Y
L = M+ 1
OD
ELSE DO
K = Y- 1
L = M+ 13
OD
FI
K = INT (365.25 * K)
L = INT (30.6 * L)
RETURN K + L + D- 621049
END DAY #
Innovations from the
Hobbyist Community
I have been reading Kilobaud
since it was first published, and I
especially enjoyed the quality of
the articles in the January issue.
No one publication can serve the
diverse needs and interests of the
rapidly growing microcomputing
/small-systems field. And Kilo-
baud is providing me and other
enthusiasts with its unique ap-
proach. Let me be specific.
Dr. Michael Wingfield’s arti-
cle, “Hardware Program Reloca-
tion,” was a breakthrough for
the small -system user. There is no
reason why many of the sophisti-
cated features that accompany
the medium and large-size com-
mercial systems cannot be devel-
oped for small systems . . . and
Dr. Wingfield’s article is more
evidence of this. Program rep-
eatability is essential if small sys-
tems are to maximize their poten-
tial power. Dr. Wingfield’s arti-
cle represents a milestone in hard-
ware development for the micro-
made of loaders including ap-
plications programs that load
data, subroutines and user com-
mands. Mr. Parsons’ fluid writ-
ing style enables him to present
complex topics in down-to-earth
language . . . which all of us ap-
preciate.
These were but two of the
many excellent articles in issue
No. 13. They represent the kind
of quality that I hope will con-
tinue to be part of your magazine
and from which others with a de-
sire to write can learn. Please pass
on my thanks to both of the
above authors and keep up the
good work.
Dr. Paul R. Poduska
Lowell MA
Objective Review
The January 1978 Kilobaud
was the best yet. The articles
“Has Godbout Done It Again?”
and “The TRS-80: how does it
stack up” are very beneficial to
those of us who are deciding on a
system but who haven’t the
knowledge or resources available
to make an intelligent decision.
I also found the article “Un-
derstanding Loaders” very infor-
mative and helpful in my coming
to a decision on a system.
Please give the readers more ar-
ticles concerning product evalua-
tion and performance vs price.
Very few of us have other re-
sources (computer stores) for
which to make these evaluations.
Your critical advice will be great-
ly appreciated by your readers.
Dwight T. Still
Lawrenceville GA
Micro Maestro Modification
I tried Terry Perdue’s “Micro
Maestro” program in the Jan-
uary 1978 issue. Given the sim-
plicity of the program and the
“interface,” I was pleasantly sur-
prised by the results. Terry must
be complimented on his arrange-
ment of “Somewhere My Love.”
After playing with it for a while, I
became frustrated by the pro-
gram’s inability to play rests. I
modified the program slightly to
allows rests— a whole rest (the
length of a note) by coding $02 in-
to the table, and a “half” rest
(one-half the length of a
coding in a $01. The change is
simple and adds only 21 10 bytes to
the code. It makes my version of
“Mary Had a Little Lamb” much
more understandable.
I also set it up so the program
gets its speed parameter from a
table (the speed value), instead of
writing it into the code. This
makes it possible to burn the pro-
gram into a PROM and still be
able to change the speed.
I have tried both these changes,
and both are incorporated in the
listing shown. As I write this let-
ter, another change, which I
haven’t tried yet, occurs to me.
As the program is written, it plays
eight different versions of the
song, then repeats the cycle. Ac-
tually, it only plays seven versions
since the eighth has all output bits
disabled and, therefore, plays
nothing. This silence is annoying
in a long song.
I suggest the following modifi-
cation to cure this. After the INC
$8004 instruction, insert these
two lines:
TST $8004
BEQ REPLAY
I think this will eliminate the
“silent song.”
Well, that’s about all. If you
haven’t tried the musical micro, I
urge you to do so. It’s neat!
Don Korte
Flint MI
0 B SB
B B SB B9B0
BBS? 0BB2
0054 BBB1
BBSS BBBI
B0D0
3030 7 F BBS?
0BB3 96 SB
0BB5 47
B B 06 97 57
0 0 08 7 F 6BB5
0 B B6 7 C 8 0B 4
0 B BE 7 3 8BBS
0611 BE 8854
0 B 1 4 DE SB
0 B 1 6 33
BB17 5 D
0018 27 EE
B B 1 8 Cl B 1
001C 27 14
BB1E Cl 02
8 B 20 27 1 2
0022 07 54
3B24 4 C
B B 25 06 54
ae27 09
B B 28 27 EA
0B2A 5 A
0B2B 26 FA
B B 20 B7 8804
BB3B 20 F 2
0 B 32 0E 52
R 8 34 7 F 6 BB 4
0837 81
0 B 38 0 1
0039 01
0B3A 89
BB3B 27 07
BB3D 2 B F 6
1 HAH HUS1C1
2 *
3 « (1 66 80 MUSIC FLAYING PROGRAM
4 •
5 • WRITTEN BY TERRY PERDUE
6 ♦ KILOBAUD JANUARY 1978 P 94
7 ♦
8 ♦ MODIFIED B Y DON KORTE
9 * ALLOWS HALF REST. NOTE CODE-01
IB ♦ ALLOWS WH0L REST. NOTE CODE-02
1 1 •
12 GRG
13 SPEED FD8
14 HSP0 RHB
15 TE HP RHB
16 HUSIC RHB
I 7 0RG
1 8 CLR
19 L DA A
20 ASR A
21 STA A
22 REPLAY CLR
23 INC
24 *
25 *
26 CON
27 LDS
28 NUN0TE LDX
29 PUL 8
3 B TST 0
31 BEQ
32 C HP 8
33 BEQ
34 C HP B
35 BEQ
36 STA 6
37 TGNLUP INC A
38 «
39 *
48 •
4 1
42 COUNT
43
44
45
46
47 *
48
49 HREST
5 B WREST
51 WAIT
52
53
54
55
56
57
L D A 8
DEX
BEQ
DEC B
BNE
STA A
BRA
LDX
CLR
NOP
NOP
NOP
DEX
BEQ
BRA
END
* 50
* B9BB
2
B
HSPD41
SPEED
HSPD
*8065
* 8BB4
*8085
• TE HP
SPEED
REPLAY
• «0 I
HREST
t * B 2
WREST
T EH P
TEHP
NUNOTE
COUNT
( 8B B 4
T0NLUP
HSPD
* 8BB4
NUNOTE
W A 1 T
INPUT SPEED PARAMETER HERE
HUSIC TABLE BEGINS He R E
CHANGE HARH0N1C QUALITY BY
BY ENABl 1NG DIFFERENT
OUTPUT BJT - >
POINT TO FIRST NOTE - 1
TIHE PER NOTE
PULL NEXT NOTE F R OH TABLE
LAST NOTE?
HALF REST?
WHOLE REST?
REGA HOLDS THE VALUE WHICH
IS SENT TO THE P1A
C0UHT1NG THESE AROUND GIVES
HA R HO N 1 C DUALITY TO THE OUTPUT
COUNT DOWN NOTE TIHE
DONE
AUDIO CYCLE N0( DONE YET
THIS CYCLE DONE.
STORE A HEW OUTPUT CODE
CANCEL OUTPUT FOR A WHILE
THEN WAIT
FINISHED
19
llTBLISHER’S
REMARKS
(from page 5)
to use than an outside service sys-
tem . . . and they have their in-
formation when they want it. If
someone calls up and wants to
know when to expect the next oil
delivery, the operator can quickly
find this information and tell
him. This would be impossible
with a service agency. Cheaper
and better, a good combination.
Although I was a bit disap-
pointed to find that Infotecs has
developed its own programming
language, even to the extent of
writing its own compiler, this
does make some sense from the
viewpoint of system integrity.
Even if some other firm gets a
dump of the programs, they
won’t make sense without the
compiler. It would take an enor-
mous amount of work to break
the programs down and use them
on anything else. It wouldn’t be
worth the trouble.
Most of the programming
credits go to company president
Edward Tolson. It’s the pro-
gramming that makes the big dif-
ference . . . and that’s a fact.
Interested dealers could do
worse than get in touch with Info-
tecs about distributing this sys-
tem. Infotecs, 1 Perimeter Road,
Manchester NH 03103.
Any Other Systems Around?
As far as I know this is the first
microcomputer system that is be-
ing produced, complete with pro-
gramming for a specific industry.
This looks to me like an excellent
approach, and I’m interested to
hear if any other firm has a simi-
lar system going.
Algorithmics
A small outfit from Massachu-
setts has been working hard to
come up with a small-business
microcomputer system, complete
with programming. The other
day I got a call from Algorithmics
with a request that I come down
to the Microcomputer Store in
Nashua (NH) and take a look at
their system.
They’ve come up with a very
sophisticated word-processing
system, using the new Seals PUP
microcomputer, a Diablo printer
and a keyboard of their own de-
sign and manufacture. This has
allowed them to design a set of
special keys for control functions
separate from the regular type-
writer and number-pad key-
boards.
The system has one of the more
comprehensive editing functions
I’ve seen. The printer is a fast job
that prints forward or backward
and has a nice clean look ... ex-
cellent for business letters. Dot-
matrix printers are all OK for
bookkeeping functions, inven-
tory dumps, etc., but when you
are preparing business letters you
have to have a clean impact
printer.
The system also does graphics
... it ground out a 73-point star
in a couple of minutes. I’ve never
seen a 73-point star before . . .
and how else would you get one
than from a computer?
This type of entrepreneurial
work is badly needed in our fledg-
ling industry. Dealers everywhere
are crying for systems, complete
with software. So far there hasn’t
been very much, and it is a
wonder that dealers have been
able to keep their heads above
water. The demand is there, so
let’s hope we see more total sys-
tems such as the Algorithmics
coming along.
Getting Newcomers Started
In adapting to computers, the
first steps are the most difficult.
Newcomers face a bewildering ar-
ray of electronic circuits, the
whole maze of programming and
its languages, all made even more
difficult with the barrier of a
foreign language: computerese.
Those of us familiar with vec-
tored interrupts and other such
nontrivial concepts enjoy jargon-
izing in our arcane dialect of
English.
However, let’s be nice to begin-
ners and try not to put them
down. But what can we really
recommend . . . by now most of
us have forgotten just how we got
to understand microcomputers
and we certainly are unable to
keep up with the flood of books.
So what can we suggest newcom-
ers do to get relief?
May I suggest you point them
to their nearest computer store
and a book called Hobby Com-
puters Are Here. All the funda-
mentals are there, written for the
beginner. It starts with a chapter
on what a computer is, then goes
on to how it counts, binary num-
bers and arithmetic. It explains
the basics of computer languages,
and then goes into the circuits
that make up a computer: gates,
flip-flops, etc. This is the ideal
first book. It costs $4.95 and, as
should be no real surprise, we
publish it.
AROUND
THE INDUSTRY
(from page 8)
The Noval Team
Let me introduce you to a few
TRS-80
FORUM
(from page 9)
130 S.(7,4)tS.(8,4):S.(13,4)sS.(23,4)
140 F.L = 50T056:S. ( L » 4 ) sN.L
150 S.(5,5)iS.(6,5)*S.(14,5)iS.(15,5)*S.(21,5):S.(22,5)
160 F.M=53T057:S. (M, 5) sN.M
170 S.(4,6)tS.(16,6):S.(19,6)sS.(20,6):S.(21,6):S.(27,6)
175 S . (28,6)
180 F.N=53T057»S. ( N , 6 ) : N . N
190 S. (3,7) iF.P=16T020»S. (P,7) sN.P
200 F.Q = 52T056iS. (Q,7) iN.Q
210 S. (2, 8)»S. (19, 8):S.( 50,8)15.(51,8)
220 S.(2,9):S.(19,9) : F . R= 2 8 TO 4 9 s S . (R,9) sN.R
225 S. ( 18, 13) s S. (25, 13)
230 S.(l,10)tS.(18,10)tS.(27,10)iS.(l,ll):S.(8,ll)iS.(9,ll)tS.(10,ll)
240 S.(18,11):S.(26,11):S.(1,12):S.(7,12):S.(11,12):S.(18,12)
250 S.(25,12):S.(l,13):S.(6,13)tS.(7,13):S.(12,13)
255 S. ( 18, 13) tS. (25, 13)
260 S.(l,14)sS.(7,14)tS.(12,14)iS.(13,14)
270 S. ( 18, 14) :S. (25, 14)
280 S.(2, 15)sS.(8,15)iS.(12,15)sS. (19,15)
290 S. (26, 15)
310 S.(2,16)tS.(9,16):S.(10,16)tS.(ll,16)tS.(19,16)
330 S. (27, 16) iS. (3, 17) :S. ( 19, 17)
340 F. A=28T037sS. (A,17 jN.AsS.(46,17)*S.(47,17):S.(48,17)
20
of the key people who keep things
moving at Noval. You can usual-
ly find Agoston (Ago) Kiss, vice-
president and general manager,
in a working position much like
that in Photo 8. Ago is involved
in both hardware and software
developments, and really put the
system through its paces while 1
was there. (I was told that he
came to this country from
Hungary in 1956. Because of cer-
tain unwelcome “visitors,” he
had to beat a hasty retreat . . .
and crossed over the border car-
rying a suitcase in one hand and
his son George in the other.
George is now software manager
at Noval and is shown standing in
Photo 2.)
Photo 9 was taken inside the
top-secret, level-3-badges-only,
research and development lab at
Noval. Lane Hauck (the Director
of R and D), seated on the right,
and Bill Blewett are deeply in-
volved in developing, and play-
ing, a new video game. (What a
life! I wonder if they have any
openings?)
I had a couple of other photos
to share with you, but my camera
started acting up. When I at-
tempted to use the strobe, the
charge for firing it off came out
of the camera into my left hand
. . . passed through my body
. . . and then returned to the
camera via my right hand. Felt so
good I just had to go back and try
it again!
Terry Sorenson, Noval’s Chief
Engineer, was one of those I
missed during my camera misad-
ventures, but he was sure helpful
during my visit, the invitation for
which came from Jerry Hansen,
the president of Noval. All I can
say about him is that he’s
definitely one of the “white-hat
good guys.” As I said earlier, it’s
a “people organization.” They
not only have a knack for making
you feel at home during a visit
. . . I’ll bet they make you feel
just as much at home when you
do business with them.
(from page 12)
related the right of personal pri-
vacy to large computers owned
and used by various agencies of
government and large private or-
ganizations. Before, the issue had
never been raised in terms of per-
sonal computers owned by indi-
viduals because such contrap-
tions did not exist. They do exist
now, and, I suppose, it will only
be a matter of time before person-
al computerists will be able to
purchase all kinds of information
about individuals in the same
manner they can buy mailing lists
today.
The possibility of blackmail by
the use of computer-compiled in-
formation is not absurd. The cur-
rently existing right-to-privacy
legislation, which will be dis-
cussed in the next Forum, can be
applied to personal computers by
changing a few words in a defini-
tion. Someday, perhaps sooner
than we think, we will face the
problem of licensing the use of
computer technology.
KB CLUB
CALENDAR
(from page 13)
H. G. Humphrey, chairman, c/o
The Micro-B Computer Store, 22
Lemon St., Truro, Cornwall TR1
2LS England.
If your Club Calendar announce-
ment contains timely, in addition
to general, information about
your club, please mail announce-
ment at least /wo months before
the date or dates mentioned in the
announcement. Please mail all
Club Calendar-related material
to: KB Club Calendar, c/o Steve
Fuller, PO Box 218, Spofford
NH 03462.
Attention, Authors!
If you write an article for Kilo-
baud, and if that article con-
tains program listings and/or
runs, please submit camera-
ready listings/runs. Use a
printer (borrow one if you
don’t own one), and please be
certain that the machine used
produces dark, clear copy
(also, try to avoid misspell-
ings). If you must use a type-
writer, and if it doesn’t have a
carbon ribbon, please make
sure that it has a relatively
new ribbon. Single-space (or,
preferably, use space and a
half if your machine has this
capability) particularly if the
listing or run is a long one.
This will make our job immea-
surably easier. Thank you.
350 F.A=51T054sS. (A, 17) :N.A
360 S. (4, 18 ) : S . (20, 18) t S . (21 , 18 ) s S * (22, 18)
370 S. (36, 18) i S. (44, 18) iS. (45, 18) : S . (50, 18)
380 S. (51 , 18) sS. (54, 18)
390 S.(5,19):S.(22,19)sS.(35,19)tS.(44,19):S.(53,19)
400 S.(6,20):S.(7,20):S.(22,20):S.(33,20):S.(43,20)
410 S. (49,20) :S. (53,20) iS. (8,21 ) tS. (9,21 )
420 F. A=24T033:S. (A,21):N.A:S. (42,21)
430 F.A=49T052:S. (A, 21 ) iN.A
440 F.A=10T012»S. { A , 22 ) :N. A
450 S. (26,22) iS. (27,22) sS. (31 ,22) »S. (32,22)
460 S. (41, 22:S. (48,22)
470 F. A=13T025:$. (A, 23) :N.A:S. (32,23)
480 S. (40,23) iS. (47,23)
490 S. (25,24) :S. (33,24) :S. (39,24) : S . (46,24)
500 S. (25,25) :F. A=34T038 : S. (A,25)»N.A:S. (45,25)
510 S. (26,26) :S. (44,26)
520 S. (27,30) sS. (33,30) * S . (46,30) iS. (47,30)
530 S. (38,26) »S. (39,26) sS. (43,26)
540 S. (27,27) iS. (28,27) tS. (28,27) *S. (39,27) :S. (43,27)
550 S. (28,28) sF.A = 40T043:S. (A, 28) : N . A
560 S. (28,29) sS. (32,29) sS. (43,29) tS. (44,29) * S . (45,29)
580 S. (27,31 ) :S. (29,31 ) :S. (34,31 ) t S . (48,31 ) rS. (49,31 )
590 S . (26,32) :S. (30,32) : S . (35,32) tS. (50,32)
600 S. (26,33) :S. (31 ,33) : S . (35,33) :S. (36,33) tS. (51,33)
610 S. (26,34) :S. (31 ,34) iS. (36,34) iS. (37,34):S. (51,34)
620 S. (25,35) t S. (31 ,35) :S. (38,35) sS. (50,35)
6 30 S. (25,36) tS. (31 ,36) tS. (39,36) i S. (49,36)
640 S. (25, 37) :S. (29,37) sS. (37,37) * S . (38,37) sS. (48,37)
650 S. (26,38) :F.A + 30T033:S. (A, 38) :N.A
660 S. (38,38) sS. (47,38 )
670 F.A=27T029:S. (A, 39) sN.A
680 F.A=33T037sS. (A, 39) sN.A
690 F. A=44T046:S. ( A, 39) sN. A
7 00 S. (30,40) iS. (31 ,40) :S. (33,40) t S. (41 ,40) tS. (46,40)
710 S.(32,41):S.(33,41):S.(41,41)sS.(46,41)
720 S. (33,42) «S. (41 .42)
730 F.A=46TD56:S. (A, 42) sN.A
740 S. (33,43) * S. (41 ,43) tS. (56,43) * S, (57,43)
750 S. (32,44)
760 F.A=41T051 :S. (A, 44) :N.A
770 S. (56,44) sS. (57,44)
780 S. (31 ,45) sS. (32,45)
790 F.A=51T056jS. (A, 45) sN.A
800 S. (31 ,46) : S . (32,46) :S. (51 ,46) sS. (52,46)
810 F.A=32T051 :S. (A, 47) :N.A
900 P.A.271," ****** CURSE YOU RED BARON ! ! ! ! "
910 P.A.936," TOM KASPER 5 JAN 78"
10000 G. 10000
21
PET’s First Report Card
an objective evaluation
Photo 1. A family portrait: PET, center, with granddaddy, Jolt, left,
and proud father, KIM, right.
Ralph Wells
Vice President Engineering
Inmarco, Inc.
7655 Sunset Blvd.
Hollywood CA 90046
A year and a half ago I wrote
an article for SCCS Inter-
face Magazine comparing the
eight personal computers I had
bought, built, designed, rede-
signed and debugged (or failed
to debug). At that time, the PET
was only a gleam in its father’s
(Chuck Pettle) eye. Now I have
one.
At the end of October my PET
arrived, three and a half months
after the usual “$800 cash-up-
front” type order that most of
the others required. Although
two weeks tardy, it had a better
record than any of the others,
except the Scamp and Tarbell,
which arrived on time. I’ve
come to accept late delivery as
a way of life for newly an-
nounced equipment, but I find
that most people entering the
exciting and mobile field of per-
sonal computing balk at it—
especially compared to today,
when you can walk into a com-
puter store and get products
from maybe a dozen suppliers
on an off-the-shelf basis. Fur-
thermore, you pay for them
next month on a credit card.
I continually hear the query,
“Is it worth it?” It is. The day
after I received my PET I took it
to a meeting of the Valley Com-
puter Club and was barraged
with similar questions from
people who already had their
own computers. How does it
stack up?
This “report card” is an at-
tempt to answer some of these
questions on an unbiased ba-
sis (I’m not selling anything). I
have personally built or bought
and modified three 8080-based,
five 6800-based, three
6502-based and one SC/MP-
based microcomputers, so the
PET has a lot to live up to.
Background
In December 1977, Com-
modore had never advertised
the PET, but the magazine ar-
ticles, television exposure and
convention displays made it a
pre-production marketing phe-
nomenon. In fact, I assume the
reader has already been ex-
posed to its fundamental
specifications. In case you
haven’t, just pick up nearly any
back copy of any computer
magazine starting last July
(e.g., Sheila Clarke’s article in
the September 1977 issue of
Kilobaud).
Right up to the present time
the big question has been: Can
Commodore produce what they
claim for the quoted price and
still make enough money to
stay in business? To get some
official answers from them, I
wrote a two-page letter and
received a one-sentence reply
that contained an honest ad-
mission to “crummy documen-
tation.” After this article was
half written, I had a chance to
chat with Chuck Pettle, but the
opinions herein expressed are
my own, derived primarily from
personal PETting and aug-
mented by the published
references and conversations
with sales representatives.
To start with, I’ll dive into the
deep end of the pool of con-
troversy and say that, in my
opinion, they’re going to make
it— and make it big! Not with
the model I received (serial 171),
but because of vertical integra-
tion and forward-thinking
management.
Setting the Stage
Let’s review some history to
get a perspective of the pros
and cons of grading the PET
#171. In a sense, this is more of
a mid-term interim report card
because the PET’S true poten-
tial has not yet been adequate-
ly documented. I’ve spent most
of my time trying to find out (the
hard way) just what I bought.
There is a gnawing feeling in
the pit of my stomach that they
are going to follow in Radio
Shack’s footsteps and not tell
me much more than I already
know.
From my point of view, the
PET is really the third product
from MOS Technology, preced-
ed by Jolt and KIM (see Photo
1). Although the Jolt is pro-
duced by Microcomputer Asso-
ciates, its debut was a result of
their synergistic relationship
with the then almost-unheard-
of MOS Technology. It was the
first microcomputer to really
take advantage of read only
memory (ROM) to reduce hard-
ware.
Of course, others, such as
Dataworks, with its 5K of oper-
ating PROM, preceded the Jolt
(and Altair) by nearly a
year— but the accent was on a
firmware operating system, not
a hardware trade-off. The 6530
mask-programmed chip, which
combined ROM, RAM, COUN-
TER and I/O, was, in my opin-
ion, almost as big a milestone
in large-scale integration (LSI)
progress as the microproces-
sor itself ... not so much as a
technological breakthrough
(competing devices had similar
technology), but as a practical
adaptation of an emerging
technology to take a giant step
22
forward on the path of pro-
gress.
Instead of needing a single
board for a Teletype port (as on
my Altair 8800), the whole Jolt
took up less than half the real
estate. The forever-drifting ad-
justments of the Altair were
replaced by a ROM/I-O, which
measured the speed of my TTY
and adapted itself! But the real
value of the TIM (or Demon)
6530 was the documentation.
Here were 1000 bytes of I/O and
times as much and used a 4-bit
word.
The obsolete formats aren’t
dead, yet. Heath took (in my
opinion) a step backward with
Octal I/O; my new Motorola
Educator II uses the single-bit
format. All in all, KIM was in-
deed a big step forward in its
time.
At the time KIM was in-
troduced, several other I/O
developments were also emerg-
ing. The highest impact devices
“PET gets As in three categories—
vertical integration, good engi-
neering and advanced technology.”
operating system available at
power-up, and documented in
such a way that its subroutines
could be (and were) used in
every program I wrote. It also
served as a workbook for learn-
ing practical usage of the 6500
code.
The Jolt had one big disad-
vantage— for practical pur-
poses: I had to have a $1000
TTY for a $300 microcomputer.
Enter the KIM
Vertical integration started
with the KIM. KIM used two
6530s to double the firmware
and utility. It preserved the TTY
I/O of the Jolt but eliminated
the total dependence on the
TTY. It had its own hexadecimal
keyboard, hexadecimal read-
out and cassette storage to re-
place punched tape. The single
board (plus power supply) KIM
outperformed three or four
boards in my Altair and Imsai.
It was the end of the octal-
binary (switch, LED, front
panel) I/O era. Toggling data
one bit at a time with lever
switches was popular in the
late 50s and early 60s. With
data in the 3-bit octal format,
the numerical readouts,
keyboards and printers of the
late 60s and early 70s became
popular. Although my PRO-
LOG preceded KIM by a couple
of years in adopting hex (hexa-
decimal), it cost nearly ten
were the full keyboard and TV
display. The pioneering laurels
for bringing the digital TV
display out of the high-priced
range (over $1000) and down to
where you and I could afford it
belong to Don Lancaster, who
literally wrote the book on the
subject. As the demand for low-
cost full alphanumeric key-
boards produced larger
volumes, the cost came down.
A third development was
also underway— BASIC. High-
level languages (including
BASIC) had been around for a
long time, but, without full
alphanumeric I/O, the com-
puter hobbyist had to work on
the bit, octal or hex level. This
meant working only in machine
language if you had the mini-
mum computer configuration,
such as KIM, Scamp, etc. It
took only a few hours of “bit-
banging” with op-code conver-
sion to realize that there had to
be something better— probably
BASIC or an alphanumeric as-
sembly language.
Even if you could afford the
extras to interface the neces-
sary keyboard and CRT (around
$1000 a few years ago), there
were other problems. On the
hardware side, you needed
memory— lots of it. You could
use RAM, and wait and wait to
load BASIC or an assembler.
Or, you could pay and pay ($425
for my ALS-8 assembler on
PROM) to get a resident assem-
bler, BASIC or both. An even
more expensive memory alter-
native was, and is, the floppy
disk, with magnetic bubble
devices warming up in the bull
pen.
On the software side, BASIC
has been evolving. Spurred by
the San Francisco community,
in general, and Tom Pittman, in
particular, the old original Dart-
mouth BASIC was first freeze-
dried to miniscule proportions
and then extended. But what is
more important is the cost of
good software. In the late 60s,
even moderate software sold
for thousands of dollars per
program, with additional hun-
dreds to adapt it to your
system. Contrast this with Tom
Pittman’s Tiny BASIC at $5,
Chuck Crayne’s 6800 Assem-
bler or Ed Smith’s Trace/Dis-
assembler in the $10-20
bracket, and the stage is set for
mass usage of computer
power. Mask-programmable
ROMs could utilize this soft-
ware at reasonable prices, but
only if high-volume sales could
amortize mask costs.
The time has come for an af-
fordable computer that does
not require the fervent learning
and application of hardware
and software skills heretofore
required of a hobbyist.
Enter the PET
The third entry from MOS
Technology (a fourth is on the
drawing boards) is another
significant step forward for its
time. At the time I paid my
deposit of $800, the closest
competition providing similar
specs cost more than twice as
much. The Radio Shack TRS-80
is squarely in competition with
Commodore’s PET, and the fac-
tor of vertical integration is like-
ly to keep the field small. Only a
few companies, such as Texas
Instruments (with their wrist-
watch and calculator mass pro-
duction-marketing technolo-
gies), have the high-priced
chips to pay the entry fees into
such a marketing race. Let’s
take a look at what vertical in-
tegration has done for the PET.
MOS Technology started as
an independently financed
splinter group from Motorola’s
6800 development program,
with associated legal problems
(now resolved). The resultant
6502 microprocessor started as
a “cheap” 6800. It uses most of
the 6800 instruction set, but is
(in my opinion) severely ham-
pered by its lack of a double-
byte accumulator. This defi-
ciency is somewhat offset by
page zero double-byte indexing
capability, which I’ve never
really been able to master.
Others have, however, and the
6502, which seemed to come
out of nowhere, burst onto the
scene in the Jolt as a showstop-
per at the 1975 WESCON show.
I personally feel that the real
innovation was the mask pro-
gramming of the MOS Technol-
ogy 6530 I/O chip. In any case,
MOS Technology was off and
running, nipping at the heels of
the well-established Intel 8080
and Motorola 6800.
Photo 2. Front-runners: TRS-80 CPU and keyboard, left. PET's CPU
and keyboard, right.
23
As the price of 8080s and
6800s fell under $30, the 6502
lost its price advantage, but it
was staying ahead in other
areas— primarily the KIM. In-
tel’s Intellec and Motorola’s Ex-
orcisor development systems
ran into thousands of dollars;
KIM was less than $300. Al-
though it didn’t do nearly as
much as the “biggies,” KIM,
with its superb documentation,
was an entry into the world of
microprocessors for the small-
er electronics manufacturer.
Until very recently, the lack
of a good, cheap assembly
language and trace has limited
my use of the 6502. The
availability of Chuck Crayne’s
assembler for use on the
Sphere 6800 and Processor
Tech’s ALS-8 for the 8080 has
diverted my attention from the
6502. My biggest disappoint-
ment with my PET is the virtual
nonexistence of the advertised
“system monitor.” It might
have filled this 6502 assembler
void.
Initial forays into a new field,
such as microprocessing, are
usually on a small scale, so the
KIM filled the bill admirably
(and still does). 6502s were
designed into new products,
and MOS Technology grew. It
added memory chips to its line,
which included character
generators as well as the 6502
family.
The Jolt and KIM were both
blockbusters when they were
announced ... but what do you
do for an encore? The Apple-ll
and Ohio Scientific Machines
had pushed the use of ROM
operating systems and hard-
ware/firmware trade-offs right
up to the state of the art for
6502. Something radically dif-
ferent was needed.
Enter Commodore
As an early front-runner in
the pocket calculator revolu-
tion, Commodore faced the
same overproduction, price
cutting and market-saturation
problems that had left a world-
wide trail of corporate corpses.
Mits was almost one of these,
and we all know what saved
them from disaster.
The microprocessor original-
ly evolved from calculator tech-
nology— the field in which MOS
Technology also started. To-
day, the calculator field is
headed in two directions: the $5
cheapy and the $600 wrist-
watch-calculator and/or the
sophisticated programmable
printing calculator with long-
term memory. How could Com-
modore compete with Tl and
others who had vertically in-
tegrated to produce everything
“in-house,” from LEDs and key-
boards to LSI chips? You
guessed it— they bought MOS
Technology. Commodore is
still in the calculator business,
but you have only to look at
their stock-market history dur-
ing the last year to see where
the action is, or isn’t.
When Commodore acquired
MOS Technology (and Chuck
Pettle), the PET was inevitable.
The pieces fell into place. The
major expense items for an in-
expensive computer were no
longer the microprocessor
chips (less than $10 in quantity)
nor the I/O chips, but rather the
I/O devices. The TV headed the
list, followed closely by the key-
board and cassette recorder.
The next generation of micro-
computers would require all of
these ... but was it practical?
There was the spectre of
Sphere. Note the marked
resemblance between the
brand new PET and my two-
year-old Sphere in Photo 3. The
resemblance is more than skin
deep. The built-in TVs and dual
keyboards are obvious; not so
apparent are the following: a
10K ROM operating system in
the Sphere (14K for the PET);
36 K RAM for Sphere (8K for
PET); PIA, dual cassette, TTY
and modem for Sphere (dual
cassette, IEEE, PIA and TV for
PET). Making allowances for
cost of RAM, PROM, etc., a
Sphere that was roughly
equivalent to my PET would
have cost about three times as
much. The problem lies in the
fact that Sphere Corp. went
broke about the same time the
PET was being announced.
The 4K PET was originally
priced at $500, which promptly
rose to $600, then to $800 for 8K
(the only model delivered, so
far). Even at $800, the question
in my mind (particularly after
shelling out the money) was,
“Can Commodore really do it?”
Judging from the reaction of
people I spoke to and the ar-
ticles I read, the consensus of
opinion was that they couldn’t.
When the promised delivery
date came and went (with the
same lame excuses I’ve heard
time and again, starting with
my first Altair) I, too, began to
wonder. As of December 1977,
Commodore was slipping even
further behind in deliveries.
Does this mean that they’re
following in Sphere’s shadow?
Will my PET become another
Sphere-like orphan— the Edsel
of personal computers? I think
not, and here’s why: PET gets
As in three categories— ver-
tical integration, good engi-
neering and advanced technol-
ogy. Let’s see how PET mea-
sures up to competition.
Vertical integration is,
perhaps, the greatest asset.
The PET combines the past ex-
perience of product develop-
ment (Jolt and KIM) with the LSI
semiconductor design and pro-
duction expertise of MOS Tech-
nology and the “offshore” sub-
assembly production and ag-
gressive marketing methods
Commodore developed for its
calculator line.
PET’s competitors have
equal or greater assets in one
or more of these three catego-
ries, but none can match the
vertical integration of Com-
modore-MOS Technology.
Radio Shack’s TRS-80 comes
closest. They have the best
mass sales setup in the world.
They also have the only foreign
supply expertise that can rival
Commodore’s. This is perhaps
the most important prere-
quisite for a cost-effective end
product.
The highly priced com-
ponents of a computer system
are: TV monitor, memory, CPU
(central processing unit), key-
board and cassette recorder.
Competition and mass produc-
tion have forced the costs of
CPU production down to a
point where, even if you’re mak-
ing your own microprocessor
(MOS Technology’s PET,
Motorola’s Educator II, etc.),
only small reductions in end-
product pricing can be realized.
Photo 3. Now-defunct Sphere, right, is very similar to PET, left. Note the combined keyboard, TV,
CPU, integral dual cassette controls and number pad. PET’s cassette is built in.
24
SP
-
. . | .
urn
TT
* U
T
If
f
n
T
JL
V 1 o
-I
A
l
o
! M
fill
s!
T
V
pj
M
ill
4
9
w
fm
n *
PI
3>
sial
1 T '
•1
5 ' ”1
1 Vi
SM
l ini
1 ;
i*:
uw
i «
T" ! 85 1
Photo 4. PET’s controversial “calculator” keyboard, with quasi-
standard key placement and conventional calculator number pad.
Note variety of graphic symbols available with shift. Lowercase is
also implemented (see text).
All the other items involve the
purchase of devices and/or
sub-assemblies made abroad.
The biggest item is the TV
monitor. Most hobbyist com-
puter manufacturers gloss over
this item with phrases like,
“Use your own television set with
adapter (not supplied).” A
reasonable frequency for your
TV set mathematically limits
the readability of characters to
16 lines, 32 characters per line,
caps only.
Most hobbyists soon find
that this limitation, plus com-
petition for time on the family
TV, leaves little choice but to
purchase a monitor. A commer-
cial TV monitor with adequate
bandwidth for lowercase, long-
line displays can cost almost
as much as the computer
(before it “grew”). In fact, I’m
using monitors that cost as
much as my PET. Inexpensive
but adequate TV electronics
come from Japan, Korea, etc.
So does another major acces-
sory— the cassette recorder.
The competition now takes
on an international flavor, and
International is Commodore’s
middle name. Most of Radio
Shack’s line of Realistic prod-
ucts are also imported, in-
cluding the TV monitor and
cassette recorder, which ac-
count for one-third of the cost
of the TRS-80 system. PET’s
keyboard is also imported
(more on that later).
Speaking from personal ex-
perience, I can say the busi-
ness of getting production
quantities of proprietary-de-
signed high-technology hard-
ware from overseas is a major
accomplishment. Delivery and
quality control require on-site
monitoring, which necessitates
a truly international organiza-
tion with established opera-
tions in the Orient.
Both Commodore and Radio
Shack can do this . . . but, can
anyone else? This is probably
the most important factor in
vertical integration — it sep-
arates the men from the
boys in low-cost, high-volume
production. It’s possible that
these two leaders could pro-
duce more cheap personal
computers in 1978 than all their
competitors combined— and
make money at it. Even with
years of calculator experience,
however, Commodore is having
overseas production delivery
problems (as of December). On
the other hand, the TRS-80 is
having problems getting its full
BASIC underway.
PET’s vertical integration in-
cludes LSI production by MOS
Technology, and when the dust
settles down, this may well be
the deciding factor. Initially,
the TRS-80 had an edge be-
cause it was designed with LSI
already in high production from
second-sourced suppliers.
MOS Technology has had to
cope with the learning-curve
problems of getting their new
LSI RAM and ROM chips into
overseas production. These
two items account for most of
the costs of the respective CPU
boards (see Photo 2).
As the learning curve pro-
gresses, the tables should turn
and give the PET a clear-cut ad-
vantage over all comers. PET’s
in-house volume base for the
6550 (4K, 5 V, static RAM) could
even make this chip a dark-
horse contender in the 4K mem-
ory field. In fact, I’m so im-
pressed with its performance
(despite four defective chips)
that I’m designing it into a
6800-based controller system.
The third factor in vertical in-
tegration is marketing. In this
area, the small (often garage-
type) computer company is go-
ing to have a very, very rough
time in the next year or two.
Radio Shack, with its massive
string of franchised outlets,
has a clear-cut advantage, and
its parent company (Tandy) is
opening a string of computer
stores.
In marketing, the PET is a
phenomenon, so far. The Com-
modore calculators survived in
a cutthroat marketplace; so
this, along with KIM, gives the
PET a solid foundation. It’s
been further augmented by
bringing in experienced person-
nel from competitors in the
field. Any newcomer will think
twice before going up against
this kind of marketing competi-
tion— the blue chips in this
game are expensive.
Another factor— the concept
of utility — sets these con-
tenders apart from their prede-
cessors. They are not aimed at
the hobbyist computer-addict
market (although the impact
will probably hit 7 on the Rich-
ter Scale).
The Keyboard,
Graphics and “Extras”
Both the PET and TRS-80
have recognized that the family
appeal requires electronic
game appeal. This makes a TV
and keyboard graphics manda-
tory and brings up the problem
of keyboard and/or joystick in-
put. Although both have graph-
ics capability, neither has a joy-
stick (as does the Dazzler or
Apple-ll). I’m sure that this will
become available in the future
since both have expansion ca-
pabilities to support a joystick.
There is a basic difference in
the use of graphics in the PET
and TRS-80. The TRS-80 splits
each character block into a
decoded matrix like the
Cromemco Dazzler, Apple, etc.
The PET goes a different route;
it gives a unique graphics sym-
bol to virtually every key on
both keyboards. This provides
a very large selection of fine-
line picture elements not
achieved through the older
techniques. It also provides
unique game-playing symbols,
such as the card characters of
hearts, clubs, spades and dia-
monds. Descending lowercase
characters (with shift) for all
alpha characters and reversing
white-on-black to black-on-
white are also provided.
All this flexibility poses
several keyboard concept and
design problems, since each
letter key must display six dif-
ferent characters. How can it
Photo 5. My Sphere’s original alpha keyboard was replaced as
shown. Note pasted editing and control labels on fronts of keys.
Specialized timing controls at far left are not standard.
25
Photo 7. All of the cassette record, playback and erase elec-
tronics are on this single small PC card.
be done economically? PET’s
solution was, of necessity, a
compromise. By using two cal-
culator-type keyboards (for
which Commodore tooling was
probably available) and chang-
ing the artwork on the anodized
caps, they got an inexpensive
(probably the cheapest in the
world) alphanumeric keyboard.
The alpha key arrangement is
only quasi-standard, but the
separate calculator numeric
keypad is standard. It is also
small enough for the cassette
mechanism to be mounted
alongside it and st I fit a
minimum-size case.
Both keysets are mounted on
the same cost-effectively de-
signed passive motherboard.
Since the keyboard matrix
plugs into the CPU with a single
cable (see Photo 2), it would be
possible to use a standard-
spaced keyboard in parallel
with, or instead of, the
calculator board.
The most commonly criti-
cized feature of the PET is the
key placement of the keyboard.
Keys are more closely spaced
than normal, the middle row
isn’t staggered, and the feel of
a calculator key isn’t the same
as that of a typewriter (it’s more
like a Teletype). I was told (by a
TRS-80 booster) that it is im-
possible to touch-type on the
PET. He was wrong; however, it
does take a relearning period,
much like going back to a stick
shift after driving an automatic
for years.
When I returned to my full
keyboards on the Sphere and
Imsai, I realized that I’ve always
used the hunt-and-peck meth-
od for number pad entry, multi-
ple key-control character and
special character entry. Unlike
touch-typing, most of my pro-
gramming is really hunt and
peck, and the PET is just about
(but not quite) as easy to use as
the Sphere (see Photo 5). A pro-
grammer friend and one of our
keypunchers both claim that
PET’s keyboard drives them up
a wall. But then, how many PET
customers are professional
data processsors?
I understand that the next
model PET will have a full key-
board, but will cost a lot more. I
could easily wire a $40 key-
board to replace the original
—in fact, the original lousy
alpha keys on the Sphere
shown i,' Photo 5 have been re-
placed, just that way. Then
what would I do about the 70
graphics and special charac-
ters that aren’t available as
standard key tops? In short,
PET’s keyboard isn’t great, but
neither are the practical alter-
natives.
Another quirk of the PET is
that its graphics and lowercase
display modes are mutually ex-
clusive. It initializes to the
graphics mode, and requires a
POKE 59468,14 to convert the
display to lowercase. A POKE
59468,12 returns to graphics.
This is accomplished by some
mysterious hardware/software
manipulations involving a PIA
and ROM that I haven’t deci-
phered, yet. You can’t mix
lowercase and graphics.
Changing modes changes
every shifted character on the
screen, but not in memory. It
can create some weird effects
that I used to change graphics
each second in an experi-
mental STOPWATCH program.
PET didn’t list its lowercase
capability in specifications at
the time I bought it, so it came
as a pleasant surprise— one of
several “extras.”
A real-time clock is another
of these extras. It doesn’t do as
much as an S-100 real-time
card, but it doesn’t cost an ex-
tra $130, either. It outputs a six-
digit, 24-hour clock word, e.g.,
Tl$ = 235959 = 23 hours, 59
minutes and 59 seconds. At
240000 it resets to 000000 and
is software presettable. It also
outputs JIFFIES, which are 1/60
second counts accumulated
from 000000. JIFFIES are about
as fast as anything you could
use with any program written in
BASIC. The clock runs off of the
8 MHz crystal.
Although it isn’t immediately
evident, the real-clock function
is an excellent example of the
aggressive design policy that
makes the PET a technical step
forward, regardless of price. I
haven’t figured out exactly how
ley did it . . . but what I’ve
^ciphered so far indicates an
impressive utilization of the
latest LSI capabilities from
MOS Technology (more vertical
integration here, and a valuable
feature not available from their
competitors). Among other
things, the Tl (time) function is
a fundamental building block in
automated home program-
ming. Since it runs on inter-
rupts, it will keep the time of
day as long as power is left on.
But, unless you trim the
oscillator, you’ll have to keep
readjusting the readout.
The Recorder System
By now it should be evident
that the PET’s low price was
not achieved by making a cut-
down, stripped version of older
technologies. Take the built-in
cassette recorder, for instance
(Photo 6). In all my other sys-
tems, built-in recorders are not
provided. Except for the TRS-80
and PET, cassette recorders
are a “hidden” extra expense of
personal computing. The gar-
den-variety cassette recorder
isn’t optimized for digital re-
cording. It sacrifices signal-to-
noise for low harmonic distor-
tion and ignores phase distor-
tion. Its electronics are an
Photo 6. Gutted cassette is probably a stopgap measure. Note
absence of usual electronics, speaker, jacks, etc.
“Except for the TRS-80 and PET, cas-
sette recorders are a hidden extra
expense of personal computing.”
26
overkill, including automatic
gain control which prevents
full-level recording.
PET’s cassette takes a radi-
cal departure. All the erase-
record-play electronics are on
the single card shown in
Photos 6 and 7. Obviously, the
gutted mechanism in the cur-
rent models is a stopgap solu-
tion to overseas delivery prob-
lems, and the eventual recorder
should be produced at a signifi-
cant savings over competing
systems.
The recording method is a
compromise between dc
saturation digital recording
and the frequency-shift-audio
techniques currently in vogue.
Dc erase is used, and square
waves are fed directly to the
record head. The record current
is limited to prevent complete
saturation and biased for
centering. On my unit this
results in about 8 db better
signal output on playback with
improved phase distortion
characteristics. My unit also
had two dry-joint solder inter-
mittents.
To find these, I had to create
a schematic. I also needed the
information to find out why my
PET played back its own tapes
flawlessly, but couldn’t copy
from one cassette to another
as I’ve been doing with my
Sphere, etc. The problem was in
the reduced record level and
phase distortion. It worked
most of the time, and might
even be practical for short pro-
grams, but it certainly isn’t
good enough for longer ones or
file storage. PET got some
demerits when I found that
several playback errors were
not caught by the double-
recording check. I’m sure that a
mass cassette duplicating
operation will eventually
duplicate digital tapes in this
format, but my copy of the first
one on the market (not Com-
modore) was a disaster.
I asked Chuck Pettle if PET
was designed that way on pur-
pose to give Commodore an
edge in the prepackaged soft-
ware field. He was surprised at
my difficulties, and assured me
that the intention was to pro-
vide a truly interchangeable for-
mat for all PET users. There is
no problem in interchange of
original recordings, only dupli-
cated copies.
I really notice the absence of
a counter on the cassette re-
corder. Unless you restrict your
tapes to two or three per side,
you wait forever for the
playback to find the right pro-
gram. In desperation, I use a
separate recorder to find the
approximate start position with
a counter and then transfer it to
the PET— a real pain. Although
the baud rate is high (1100
baud), a long preamble, double-
buffered recording scheme and
a motor stop between files s^w
down the file handling tc a
snail’s pace, compared to a
Tarbell. The second cassette
port is fully implemented on the
CPU but, as yet, no recorder is
available to make use of it. I
hope it will have a counter.
Another nice added “extra”
is the verify mode. After record-
ing, you can rewind and verify
the tape playback against
memory. Since I’ve eliminated
the intermittents in the record-
er, it’s a bit redundant because
there has never been a play-
back error.
Another extra is the unri-
valed simplicity of loading a
program— turn on power, insert
a cassette and press RUN. It
tells you to play the recorder,
displays the label of the first
thing it finds, tells you it’s
loading and if it loaded OK and
runs the program. Even a very
small child can do it. An A
rating. If children are to realize
the maximum educational po-
tential of personal computing,
this approach will be very help-
ful. If you specify a label, it will
display each label it finds until
it gets the right one— then
loads it.
The PET’s recording format
is unique, like those of most of
the new computers ... it looks
as though the Kansas City
standard will bite the dust. The
PET maximizes the hard-
ware/software trade-off. It uses
almost a bare minimum of
analog devices (room for de-
sign improvement here), a cou-
ple of PIA ports and no UARTs
or other serial I/Os. It’s the
most cost-effective digital
recording system I’ve analyzed,
although the Educator II is a
close second. It’s an A + exam-
ple of saving money with de-
sign ingenuity.
The TV Board
All of the other competing
computers with CRTs use off-
the-shelf monitors or TV modi-
fications. In this instance, as
with the recorder, PET breaks
with tradition, gaining im-
proved cost/performance by re-
placing hardware with firm-
ware. The complex sync sig-
nals, which use up hardware in
both the traditional character
generators and monitor, are
generated by firmware and the
very powerful 6522 I/O chip. The
video, horizontal and vertical
drives are also available on the
rear user terminal. Because the
video board doesn’t need to
decode sync o- amplify video,
it’s simpler (anu cheaper) than
competing models. Since the
screen is built in close to the
operator’s eyes, it can be
smaller than a separate mon-
itor (such as the TRS-80) and
still provide the same legibility
—another saving.
There is only one external ad-
justment: contrast. My PET
needed vertical centering. It
was done with the black tabs
on the neck of the CRT. A small
pot at the rear adjusts the
height. So far it las been very
stable and provides a steady
picture with a superior band-
width, another A-rated example
of cutting costs with creative
system design.
The CPU Board
Photo 8 shows the CPU
board— PET’s brain. It takes
less than two minutes to
remove it. Wiring harnesses
cost money. Both the PET and
the TRS-80 keep them to a
minimum. The board plug-
connects to the power upply,
keyboard, video and recorder.
Incidentally, be careful with the
keyboard plug. Mine became in-
termittent after its first replace-
ment. The leaf spring contacts
in the female cord connector
are easily overstressed and
Photo 8. PET’s brain: Top 16 chips are RAM. Seven ROM chips
below contain operating firmware. Power supply and cassette #7
are along right side. Output ports are along bottom (rear). BUS and
memory expansion are at left.
27
Photo 9. The power transformer, filter capacitor and 110 ac control
are the only electrical devices directly wired to the chassis.
may have to be re-formed with a
probe.
Note how the four expansion
connectors are made directly
to the board through slots in
the side and rear of the case— a
far more efficient arrangement
than that of any of my other
systems. At this time there is
nothing available to connect to
them, but when there is, the dif-
ference between the utility of
the PET and the TRS-80 is likely
to give the PET a big com-
petitive edge (see Photo 2).
The long connector on the
left-hand side has what the
TRS-80 has on its single expan-
sion port. In addition, the ad-
dressing is available decoded
into 4K blocks. Current plans
call for its use in RAM, ROM
and PROM (2716) expansion.
The monitor and assembly lan-
guage will probably go into
ROM.
The current price of $200 for
4K of RAM makes PET about
the highest-priced RAM on the
market. When the 6550 moves
out on its learning curve, PET
should be in a position to pro-
vide the cheapest memory
around.
The small connector pad in
the lower left corner is for
cassette #2. You can play the
recorder into it. It works, but, as
yet, there isn’t any recorder
available to use with it. If PET
doesn’t make one available
soon, I’m sure someone else
will, and I hope they provide a
counter. The center connector
brings out the aforementioned
video feeds and half of the
powerful 6522 PIA program-
mable I/O. It’s called a User port
and, if documented adequately,
could become PET’S most
valuable asset.
The lower right connector is
the IEEE-488 bus. If and/or
when the S-100 bus system
yields to another format, it’s
likely to be the 488. This system
is supposed to allow your PET
to talk with up to 18 peripherals
through a high-speed, 8-bit
parallel bus. Properly im-
plemented, it can be almost as
fast as a motherboard or back-
plane.
There are more than 200
devices (a lot from Hewlett-
Packard) available for use with
the 488. However, most of them
cost more than the PET and are
special-purpose test instru-
ments, not really suited to per-
sonal computing. Motorola and
others are coming out with LSI
chips that should make the 488
system cost competitive with
the S-100. This won’t happen
immediately, but when it does,
PET will have a well-
established lead over the rest
of the pack, particularly in soft-
ware. PET gets an A here be-
cause Commodore’s vertical in-
tegration should allow them to
make inexpensive peripherals
that could be used with com-
petitors’ microcomputers, as
well as with the PET. The
TRS-80 (see Photo 2) with its
single, unique 40-pin port only
rates a D when it comes to this
kind of expansion.
PET’s power supply, see
Photo 9, is 5 volts only (Sphere
uses five different voltages) for
the digital equipment. The TV
board has its own rectifier-reg-
ulators. The CPU board splits
the load into three sections
with the three 5-volt regulators
along the left-hand side. The
two power transistors with heat
sinks are the motor controllers
for the cassette recorders. The
regulators are running hot now,
so additional loads should be
limited.
The 8 MHz crystal clock
drives the 6502 microprocessor
at 1 MHz. It also provides the TV
timing and 60 Hz JIFFIES. The
crystal is stable, but the factory
feels that plus or minus IV 2
minute per day is adequate. If
you want greater accuracy,
you’ll have to trim the driving
capacitors next to the crystal.
A 6-30 pf variable in parallel
with 22 pf did the job for me (see
Photo 10). Now I can trim it like
my digital wristwatch. The
24-hour clock is counted with
interrupts and should be soft-
ware independent. I’ve en-
countered unresolved prob-
lems with a program that con-
tinuously reads Tl$ — it speeds
up the displayed time.
The 6550 RAMs are 4K, high
speed, low power, static, and
require only 5 volts. They are
pinned as IK by 4 bits, so they
are socket-mounted in pairs
along the front of the board.
Page 0 is at the left and the high
nibble is toward the front. If
memory problems occur (I’ve
had four failures), you’ll need to
play musical chairs, since it’s
impractical to apply a memory
test to the low IK where BASIC
operates its scratchpad. This
device gets an A for design and
a D for deportment.
The ROMs in the first units
(mine included) were not the
MOS Technology devices cur-
rently being shipped. They are
2K devices and are now being
soldered in. Although PET is of-
ficially specified for 14K of
ROM, 2K of the same ROM is
used as a character generator.
The PET is currently oriented
toward the personal-computer
mass market; changing only
the ROMs and keyboard caps
could make it a super develop-
ment system, smart terminal,
dedicated controller, word pro-
cessor, typesetter or just about
anything micros are, or will be,
used for. It could happen vir-
tually overnight, and, with the
inherent mass-production
economics, it would be a price-
cutter in any market. (That’s
awesome when you think about
it, since MOS Technology could
supply inexpensive masked
ROM for any application.)
Mechanical Engineering
PET gets a B+ for its metal
case. It will probably be re-
placed by a more durable plas-
tic case, but dies for this size
molding are a long time com-
ing. In either case, the PET is
utilitarian and its exterior ap-
pearance can only be com-
pared to units costing several
times more. It even has a prop
to hold up the hinged top for
servicing. The tooling is a little
sloppy and some of the holes
are mismatched. My degree
was in mechanical engineering
(a long time ago), and I ap-
preciate good mechanical de-
sign. PET has it. Not only is the
case impressive, but so are the
circuit-board layout and the
overall cost-effective design
decisions. Three of the four cir-
cuit boards are inexpensive
“single sided.”
The case of the TRS-80 is a
good design job also, but the
overall effect looks like a key-
board with dangling wires to a
dominating TV, with a cassette
and power supply strung
28
around it. The TRS-80 is more
attractive than the uncased
Jolt or KIM, but, to the average
neophyte, it may not look like a
computer when compared to a
PET or Sphere.
And Now . . .
The Bad News
PET gets low marks in two
areas: reliability and service
maintenance. I give it a D. At
the same time, there is enough
room for improvement so that it
could go to the head of the
class. It worked when I received
it. Since then, I’ve had four in-
termittents; three were bad
solder joints and the fourth was
a defective connector. I have
also had four memory failures,
a glitch in my TV horizontal
sweep, drifting vertical center-
ing, undetected read errors, off-
frequency crystal calibration
and a couple of other weird
goings-on that remain uniden-
tified. To put things into per-
spective, I should add that this
behavior is better than that of
my Mits 8800, Mits 680, Imsai,
Sphere, Jolt or SWTP.
Bugs are a way of life when
you get the first units off a pro-
duction line; I expect them.
Mits had trouble with bad mem-
ory chips on the first 8800
boards . . . worse than my
PET’s. They wouldn’t send
replacement ICs so I reluctant-
ly sent the useless boards
back. It was four months and
$40 extra before I got working
memories from them.
The big hang-up with bugs in
my PET is that there is no ser-
vice information provided; fur-
thermore, it’s unlikely that I’ll
see a schematic for a long, long
time, if ever. The local dis-
tributor doesn’t have any more
information or spare parts than
I do. The 6550s aren’t on the
market and there are no com-
plete spec sheets available
for them. A magazine article
had estimated that factory ser-
vice would require two months,
including shipping. If you
detect a note of frustration,
you’re right! It’s even worse
when you see a little LED on the
board and know that it’s a part
of a built-in diagnostic system
that’s using up some of the
ROM you bought. Neither you
nor your local dealer can use it;
it’s a factory secret. Now what
do you do?
First, call the factory. When I
called, the girl who answered
didn’t know what I was talking
about, and the fellow who
might have known was unavail-
able. People who went through
this with Mits and SWTP in the
“old” days (it’s changed now)
know the script.
After a period of fuming and
fretting, punctuated with ex-
pletives, I decided that $10,000
worth of test equipment and
four years’ worth of experience
with microprocessors ought to
be able to solve the problem
without schematics. It did
—partially.
I had to write my own mem-
ory test program and use a
multitrace storage scope to
eventually find the intermit-
tents and some of the bad
memory chips (also intermit-
tent). Then, another call to the
factory. This time I was put
through to the right man with
the right attitude and right
answers— a real gem. Three
days later I had replacements
and spares, no extra charges,
no insistence that I relinquish
my cherished PET for an in-
definite stay and a lot of good
solid advice on how to tackle
the remaining problems. He
also assured me, as did Chuck
Pettle, that most warranty
repairs took less than a week, if
worse came to worse.
OK, so my PET is running
pretty well, but what about the
housewife in some boondocks
town without a well-equipped
laboratory, years of experience
or a WATS line? What if she got
my #171? Well, as of December,
her only recourse would have
been to return it to California or
Pennsylvania and hope that
Murphy’s Law, as applied to in-
termittents, wouldn’t require
too many return trips. However,
by the time you read this, PET
could be in the best service
position any personal com-
puter manufacturer has ever
been in.
The information and special
wiring harness should be
released so that the built-in
diagnostics can be utilized by
relatively inexperienced peo-
ple. Faults could be fixed by
identifying and exchanging the
offending circuit. Since there
are only four circuit boards and
a rudimentary power supply,
the built-in diagnostics, aug-
mented by test cassettes,
should easily bracket the
problem.
From personal experience,
I’d estimate that most “while-u-
wait-repairs” could be done in
less than 15 minutes. The abili-
ty to do this was obviously a de-
Photo 10. Author’s modification of 8 MHz crystal oscillator with
trimmer capacitor trims 24-hour clock to high precision, but soft-
ware problems remain.
sign objective. Currently, there
are two flaws in the grand plan.
All available parts are being
used to try to satisfy a huge
backlog of delinquent system
orders. There are no spare
boards for dealers or service-
men. Also, documentation and
test equipment are not yet
available in what Chuck Pettle
describes as an “acceptable”
form.
When I asked him when I
would get schematics ade-
quate for servicing the prob-
lems with my PET, he told me
that only the characteristics of
the I/O were going to be re-
leased, and the rest would be
kept “secret from competi-
tors.” In a vain attempt to get
him to change his mind, I
pointed out that a competent
computer engineer could pro-
duce a schematic of the whole
system in a few days and that
any programmer who has writ-
ten a BASIC interpreter (see “A
Tale of Four BASICS,” Kilobaud
No. 13, January 1978) could pro-
duce a source listing of the
ROMs. In fact, the only firms
that possess these in-house
skills are his competitors'. As
they say about gun control, “If
you make gun possession a
crime, then only criminals will
possess guns.” If PET (or Radio
Shack) refuses to supply sche-
matics to servicemen and pro-
duct designers, then the only
people who can get the infor-
mation are their competitors
with skilled manpower.
I admit I’m biased by the
many wasted hours l‘ve spent
debugging my PET, but I can’t
help feeling that Chuck is
adhering to a shortsighted
policy. However, I feel that he’s
a reasonable man, so I hope
someone else will succeed
where I failed, and we’ll all
benefit.
After rereading what I’ve just
written, it’s evident that, with
the exception of the service
and documentation problems
(which may not exist by the
time this is printed), the PET
has been depicted rather posi-
tively. As a matter of fact, Com-
modore could easily drop a per-
haps fatal wad on the PET ven-
ture. Several local dealers who
29
were pushing PET a month ago
are now telling customers to
buy something else because
“Commodore is going broke.” I
suspect that delinquent deliv-
eries and “cash-up-front”
dealer policies are the real
motivation, but how much of
this can PET take?
One look at the gutted cas-
sette recorder implies a big
problem with overseas sup-
plies. Less obvious, but un-
mistakable, evidences abound
to attest to the probability that
my cold-solder-joint intermit-
tents are the result of question-
able production practices and
relaxed, or inadequate, quality
control.
No matter how cost-effective
a product design may be or how
dynamic the pre-production
sales effort, if you can’t pro-
duce a reliable product on
schedule with efficient and
minimal after-sales service,
you’ll lose the ball game ... re-
member Viatron? MOS Tech-
nology had problems with the
early KIMs (mine went back
twice), and successfully solved
them. I’m betting that PET will
have a similar success.
When it comes to software,
PET gets a C, with an “in-
complete” noted in the margin.
The bare-bones listing of Micro-
soft’s latest BASIC makes it dif-
ficult to work with, much less
evaluate. Someone else will
have to do that after the manual
is published. So far it’s about
the same as the Crayne BASIC
I’ve been using on the Sphere
and the Mits on the Altair. It’s
faster, the error messages are
better and the files are double
buffered, but watch out for
commas within quotation
marks, such as addresses in
FILE programs— they tend to
act as delimiters. If you’re used
to using abbreviated instruc-
tion, you’ll be disappointed.
The original specs called for
a 4K basic operating system.
Compared to my Sphere operat-
ing system with only 2K of
PROM, the PET is a disappoint-
ment. There are USR and SYS
commands in BASIC, but no
facility to load or generate
machine code except by writing
your own program to POKE it in
BASIC. I had hoped that they
would at least start where the
two-year-old Sphere system left
off.
If I were to put the Crayne’s
Sphere BASIC in ROM along
with the current ROM operating
system that consists of V3D,
PDS, Mason’s X-DBUG and Pro-
gramma Assoc, text editor, it
would require 20 percent less
ROM and provide many fea-
tures not found in this version
of the PET. This includes utility
subroutines such a number-
base conversion, multibyte divi-
sion and multiplication, block
moves, hex-decimal-ASCII con-
versions, etc.
Conclusion
After all the pros and cons
have been considered, it looks
to me as though Commodore’s
PET has the brightest future of
any microcomputer I’ve ever
evaluated. It could graduate
summa cum laude. Right now
it’s on shaky ground and could
conceivably flunk out, as did
the Sphere. It could have the
short-term success of the
average microcomputer, such
as the Jolt. No matter how
history marks its final report
card, a new era of mass usage
of artificial intelligence has
been ushered in by Com-
modore’s PET.B
“If PET refuses to supply
schematics to servicemen and
designers, only competitors
can get the information. ’’
(""Products that make your computer useful”'!
EXTEND... CONTROL!
I
L
I A
Whether for troubleshooting or analysis, if you have an
S-100 machine at some point you will need our Extender
Board with Logic Probe Kit ($35). The logic probe makes it
easy to see which signals are going where . . . our special
edge connector provides easy clip lead probing, jumper links
in supply lines allow for fusing/current measurement/shut-
down independent of system, and a non-skid needlepoint
probe helps prevent accidental shorting. As with other
Mullen kits, you also have quality parts, detailed instructions,
and a realistic price.
The Altair/S-100 compatible Relay/Opto-Isolator Control
Board Kit ($117) is a natural for controlling audio systems,
dme lapse photography experiments, model trains, robot
devices, or any application where you need a number of in-
telligent switches . . . more uses are discovered daily, as
detailed in our applications notes. 8 reed relays respond to an
8 bit word from your computer; 8 opto-isolators accept an 8
bit word from the outside world and send it back to your
machine for handshaking or further control purposes. In-
cludes detailed instructions.
MULLEN COMPUTER
board's!
BOX 6214, HAYWARD, CA 94545
M32
Available by direct mail
(shipped ppd. in USA from
B stock; Cal res add tax) or at _
| many fine computer stores. I
^Dealer inquiries invited. I
30
SYSTEM 9710
I SELECTERM
TheSELECTRICII* Printer
you can TRUST
CAUSE ■ ■ alt's brand new,
and fully assembled and tested.
BECAUSE . . ■ After extensive engineering design and testing by Micro
Computer Devices, IBM Corporation has approved the SELECTERM for use with
your microcomputer, and provides you with their factory warranty and yearly
service agreement for the typewriter. In addition, the electronics conversion
portion is fully warranteed by Micro Computer Devices.
BECAUSE . . ■ You can connect the SELECTERM to your computer within
minutes of taking it out of the carton.
IT S THAT EASY! AND THAT RELIABLE!
FEATURES
■ Complete ASCII character set
in supplied element.
■ Full upper, lower case alpha-
numeric characters.
■ Tab Command, Index (verti-
cal tab), Backspace, Bell— all
under computer control.
■ Parallel Interface, standard.
ALL ELECTRONICS INCLUDED
■ Power supply, electronics and
cable sets included to permit
immediate connection to the
parallel port of any computer,
at standard TTL level.
SOFTWARE
■ All necessary conversion soft-
ware in PROM to handle
ASCII input, directly.
PRINTER or TYPEWRITER
■ May be used as a standard
typewriter when not in use
with your computer.
OPTIONS
■ Dual Pitch, $125
■ Correction Feature, $125
■ Tractor Feed Platen, $250
■ Noise Reduction Feature, $50
PRICE and DELIVERY
■ Assembled and tested, $1750
■ Available ONLY from author-
ized dealers.
■ Delivery 1 to 2 weeks from
receipt of order.
■ OEM delivery in quantity within
30 days.
AVAILABLE SOON
■ RS-232 Interface
micro
computer
devices
inc.
960 E. Orangethorpe, Bldg. F
Anaheim, California 92801
Telephone (714) 992-2270
* Registered trademark of IBM Corporation M30
" Innovators to the Microcomputer Industry "
31
Bob Buckman
3954 Hillview
Santa Maria CA 93454
Scope Power!
a review of Tektronix’s Model 922
T ektronix’s motto is “Com-
mitted to Excellence.’’ That
motto, and the resulting equip-
ment, prices most of Tektron-
ix’s products out of the hob-
byist range. In 1976, Tektronix
announced the T900 series of
oscilloscopes. Finally, a Tek
scope I could afford! Last June
I purchased a medium-priced
T922.
You software types should
know what an oscilloscope is.
It can display little squiggly
lines to enable you, or your
friendly neighborhood service
center, to troubleshoot your
hardware. Read on for a short
course in scopes.
Tektronix has always been
the Cadillac of the industry,
with appropriate prices. The
great thing about the 900 series
is that the basic design is the
same as for the top-of-the-line
scopes. Why should you buy
this scope? What features put
the 900 series above all the
others in the market?
First, the scope tube itself.
Designed and constructed by
Tektronix, it contains no manu-
facturing shortcuts. The scope
is literally built around the
scope tube. The huge 8 x 10
centimeter square screen with
internal graticule on the front
of the display won’t rotate out
of place, rub off or fade with
age. With the nominal 12,400
volts dc acceleration potential,
the dot size is small and the
writing speed fast.
The least expensive scope
in the 900 line is a single-trace
instrument that sells for $650.
All the others in the line are the
dual-trace variety (and only
Photo 1. The complete system.
slightly more expensive).
The scope I bought is the T922
—a dual-trace, 15 MHz band-
width, portable instrument that
tips the scales at a mere 15
pounds (see Photo 1). I ordered
it from stock by telephone from
my local Tektronix Service
Center and mailed a check; two
weeks later, UPS left it on my
doorstep. Ah! Nothing beats
the joy of opening a box with
Tektronix printed on the side! It
was packed securely with air
space all around and contained
a manual and two $42 list-price
X10 attenuation probes (includ-
ed in the price— $850 FOB Bea-
verton OR — you thought I
wasn’t going to tell you). It
worked perfectly; that’s one ad-
vantage of buying an assem-
bled instrument.
I was surprised at the length
—almost 19 inches, most of
which is scope tube ... I
couldn’t resist looking inside.
The plastic case comes apart
after removal of six bolts (see
Photo 2). Most of the acton oc-
curs on two single-sided circuit
boards. The pilot light is a neon
NE-2 with a light pipe guiding
the way to the front panel. Most
knobs are extended with plas-
tic rods to controls positioned
toward the rear of the circuit
boards. The attenuators, since
they are of unique stripline de-
sign, are mounted on the front
panel near the BNC input con-
nectors— altogether an easily
assembled, well-planned lay-
out, with room for expansion
(see Photo 3). I may add the
T935 delayed sweep features
as soon as I get the other scope
manual.
The 15-position calibrated at-
tenuators are constructed us-
ing the same stripline cam-
switch techniques from the 500
MHz mainframe machines.
With steps from 2 millivolts to
10 volts and a variable control
over a 2.5 to 1 range, any volt-
age can be easily displayed.
The 912 and 922 calibrated
time base has 20 steps in a 1-2-5
sequence from 0.5 second to
0.2 microsecond per centime-
ter. With a variable control from
IX to 10X, the maximum sweep
rate is 20 nanoseconds per
centimeter!
32
Photo 2. The tube determines the scope’s length,
width and height (right side view from front).
Photo 3. Most controls are extended (left side
from front).
Frequency response? Oh
yeah— dc to 15 MHz for the 912
and 922 (my scope), dc to 35
MHz for the 930 series. That’s
minimum, folks! The top trace
in Photo 4 is an 8080 phase 1
clock; the bottom trace is
phase 2. The slight ringing on
the low side of the phase 2
clock is due to a bad ground to
the scope probe (see Photo 4).
Triggered sweep assures
that the sweep does not start
until the triggering conditions
are met. This allows you to easi-
ly measure single or repetitive
pulses, which will always show
up at the same place on the
screen. You can select the posi-
tive or negative slope of the
waveform and, by varying the
trigger level, trigger anyplace
on the waveform. Auto trigger-
ing, alternate or chop mode
(dependent on sweep speed),
TV field or line sync, external
sweep and X-Y modes give you
any combination of triggering
modes you need.
Ever lose the trace? You
know the signal is there but the
dc level has moved it off the
screen somewhere. Simply de-
press the beam-finder button.
The display will be squeezed
vertically and horizontally so it
will fit on the screen no matter
where the position controls are
set. Once you’ve found the
trace, set the controls for best
position and release the mo-
mentary beam-finder button.
The balanced delay line (the
looped cable in the center of
Photo 2) slows down the input
signal so the sweep starts be-
fore the display— you can see
the part of the signal that
started the sweep. No more
guessing what the leading
edge looks like!
The manual provided with
the scope is a work of art. It’s
actually a scope textbook all by
itself: 20 pages of operating in-
structions, five pages of per-
formance tests to let you know
if the scope still meets its spec-
ifications, 31 pages of service
information to help you fix
whatever might be wrong, at
least 50 pages of fold-out block
diagrams and detailed sche-
matics, complete exploded
views of all mechanical parts
with a detailed parts list for
everything — truly a joy to
peruse.
Did I say anything about re-
sale value and reputation? Ten
years from now, I could sell this
scope for nearly the same price
I paid for it. Tektronix’s world-
wide reputation is simply the
standard of excellence.
This report has been flatter-
ing. What didn’t I like? Only two
things bothered me. First, the
chop and alternate sweep
mode is selected internally by
the sweep-speed switch in the
T922, but the rack-mount ver-
sion of the same scope has a
front-panel switch! I know that
the chop mode isn’t very useful
above half a millisecond per
centimeter, but sometimes it is
essential for single-shot events
and I can’t get to it. Frustrating.
The second thing is a feature
you don’t really think about un-
til it is time to take a scope
photo: a scale illumination. It
would require front-panel rede-
sign and add at least a hundred
bucks to the price, so I accept
the lack of scale lighting. The
rack-mount version does have
scale illumination and takes
any 7000 series scope camera
—it also costs more.
If you’re deciding on a scope,
consider the Tektronix 900
series. Why not the best? I
know my scope is one of my
best investments. A card to
Tektronix, PO Box 500, Beaver-
ton OR 97077, or a call to your
local Tektronix representative
will result in your getting order-
ing information.*
Photo 4. 8080 clocks 200 nanoseconds per division.
33
ALABAMA
Huntsville
Computer Land
ARIZONA
KANSAS
Phoenix
Bits & Bytes
Mission Computer Center - Byte Shop No. 61
Tempe
Byte Shop
Wichita
Computer System Design
Yuma
Ozymandias Systems
KENTUCKY
CALIFORNIA
Louisville
Cybertronics
Santa Ana Advanced Microcomputer Products
Louisville
Data Domain
Costa Mesa
Algorithm Personal Computers
LOUISIANA
Hawthorne
Applied Process Laboratories
Downsville
Bill Gulledge
San Rafael
Aaron Enterprises
MARYLAND
Long Beach
Avido Electronics
Towson
Computers, Etc.
Fullerton
Bits N Bytes
Rockville
The Computer Workshop
Berkeley
Byte Shop
MASSACHUSETTS
Burbank
Byte Shop
Boston % American Used Computer Corporation
Fresno
Byte Shop
Waltham
The Computer Mart, Incorporated
Haywood
Byte Shop
MICHIGAN
Lawndale
Byte Shop
Berrien Springs
The Abacus
Palo Alto
Byte Shop
Royal Oak
Computer Mart, Incorporated
Pasadena
Byte Shop
Livonia
GAW Computertronics
Placentia
Byte
Grand Rapids
Jepsan
Santa Barbara
Byte Shop
Brighton
The General Computer Company
Santa Clara
Bytel Shop
MINNESOTA
San Diego
Byte Shop
Edina
Computer Depot, Incorporated
San Jose
Byte
NEBRASKA
San Jose
Byte Shop
Omaha
Omaha Computer Store
San Rafael
Byte Shop
NEW JERSEY
Tarzana
Byte Shop
Trenton
Ace Electric
Walnut Creek
Byte Shop
Succasunna
Computer Hut
Westminster
Byte Shop
Iselin Computer Mart of New Jersey, Inc.
Los Angeles
Richard Chew
Ramsey
Dollar Planning, Incorporated
Costa Mesa
Computer Center
Hoboken
Hoboken Computer Works
Van Nuys
Computer Components, Inc.
Clark
S-100, Incorporated
San Diego
Computerland
NEW YORK
San Leandro
Computerland Corp.
Levitown
Byte Shop - East
El Cajon
Computer Metrics, Inc.
Fayetteville
Computer Enterprises
Westminster
Computer Playground
New York
Computer Mart of New York, Inc.
Lawndale
Computer Stop
Manhasset
Computer Microsystem
San Francisco
Computer Store of San Francisco
Hollis
Synchro-Sound Enterprises
Hayward
Computer Systems Unlimited
Rochester
The Memory Merchants
Indio
Dean's Music City
NORTH CAROLINA
Reseda
Dynatron International Company
Boone Alpha Digital Systems, Incorporated
Fresno
Electric Brain
Raleigh
Byte Shop
Riverside
Electronic Supply
Ashville
Computer Sharing .Incorporated
Torrance
Futra Company
Kinston
Professional Computer Associates
Lawndale
Jade Computer Products
OHIO
Santa Ana
L. Electronics
Cincinnati
Cincinnati Computer Store
Anaheim
Micro Computer Center
Kent
Ohio Microsystems
Santa Monica
Mission Control
Cleveland
Tec Mar, Incorporated
Sherman Oaks
Peoples Computer Shop
OREGON
Redondo Beach
Personal Computer Center
Beaverton
Byte Shop - Beaverton
Torrance
Randal Data Systems
Coburg
Forethought Products
Bakersfield
R & H Electronics
Aloha
Kent's Komputers
Scotts Valley
Success Systems
Eugene The Real Oregon Computer Co., Inc.
San Gabriel
Sunny Sounds
Carson
Sunshine Computer Co.
PENNSYLVANIA
Tarzana
Tech-Mart
King of Prussia
Computer Mart of Pennsylvania
San Diego
The Computer Center
Huntingdon Valley Marketline Systems
Orange
The Computer Mart
Pittsburgh
The Electronics Place
Santa Monica
The Computer Store
SOUTH CAROLINA
West Lake Village
Vector Graphics, Inc.
Columbia
The Byte Shop No. 32
COLORADO
TEXAS
Englewood
Byte
Dallas Altair Computer Center of North Texas
Boulder
Byte Shop
Houston
Andy Electronics Company, Inc.
Denver
Computer Technology
Austin
Austin Microproducts
DELAWARE
Austin
Balcones Computer Corporation
Newark
Delaware Microsystems
Houston
Computerland
FLORIDA
El Paso
Computer Terminal
Leesburg
Delta Electronics
Houston
Electronic Specialty Company
Tallahassee
Florida State University
Houston
Houston Computer Mart
GEORGIA
Corpus Christi
Microsystem Services, Inc.
N.E. Atlanta
Byte Shop
Forth Worth
Tandy Computers
Atlanta
The Computer System Center
Richardson
The Microstore
HAWAII
Houston
The Computer Store
Honolulu
Mahalo Microsystems, LTD
College Station
Young Electronic Service
ILLINOIS
TENNESSEE
St. Posen
Bits & Bytes Computer Store
Nashville
Computer World
Champaign
Champaign Computer Company
UTAH
Harwood Heights
DMA, Incorporated
Provo
Computers & Stuff of Utah
Evanston
Itty Bitty Machine Company
VIRGINIA
Skokie
Lillipute Computer Mart
McLean
Computer Systems Store
Lombard Midwest Microcomputers, Incorporated
Newport News
The Home Computer Center
Oak Park
The Computer Store
WASHINGTON
INDIANA
Seattle
The Retail Computer Store
South Bend
Audio Specialists
WEST VIRGINIA
Bloomington
Data Domain
Morgantown The Computer Corner Incorporated
Indianapolis
The Data Group, Incorporated
WISCONSIN
Indianapolis
The Home Computer Center
Sheboygan Falls
DMA Incorporated
34
TARBELL SETS STANDARDS
For Hobbyists and Systems Developers
Sales to thousands of hobbyists over the past two years have proven the Tarbell Cassette
Interface to be a microcomputer industry standard. Tarbell Electronics continues research and
development to produce new and efficient components to fill hobbyists’ changing needs.
TARBELL
CASSETTE INTERFACE
TARBELL FLOPPY DISC
INTERFACE
• Plugs directly into your IMSAI or
ALT AIR* and handles up to 4
standard single drives in daisy-
chain.
• Operates at standard 250K bits
per second on normal disc format
capacity of 256K bytes.
• Works with modified CP/M*
Operating System and BASIC-E
Compiler.
• Hardware includes 4 extra IC
slots, built-in phantom bootstrap
and on-board crystal clock. Uses
WD 1771 LSI Chip.
• Full 6-month warranty and exten-
sive documentation.
• PRICE:
Kit $190 Assembled $265
Plugs directly into your IMSAI or ALTAIR*
Fastest transfer rate: 187 (standard) to 540 bytes/ second
Extremely Reliable— Phase encoded (self-clocking)
4 Extra Status Lines, 4 Extra Control Lines
37-page manual included
Device Code Selectable by DIP-switch
Capable of Generating Kansas City tapes also
No modification required on audio cassette recorder
Complete kit $120, Assembled $175, Manual $4
Full 6 month warranty on kit and assembled units
CP/M with BASIC-E Compatible Disc Drives
and manuals: $100 Ask about our disc drives priced as low as $525
• Gold plated edge pins
• Takes 33 14-pin ICs or
• Mix 40-pin, 18-pin, 16-pin and
14-pin ICs
• Location for 5 volt regulator
• Suitable for solder and wire wrap
• ALTAIR/IMSAI compatible
UllllilJlumil JII 1 1 Illllill IlIIUlliimiimiLr Price: $28.00
For fast, off the shelf delivery, all Tarbell Electronics products may be purchased from computer store dealers
across the country. Or write Tarbell Electronics direct for complete information.
TARBELL
PROTOTYPE
BOARD
Model 1010
'ALTAIR is a trademark/tradename of MITS, Inc.
CP/M is a trademark/tradename of Digital Research
20620 South Leapwood Avenue, Suite P
Carson, California 90746
(213) 538-4251 m
35
J. Tom Badgett
1917 Washington Street
Bluefield WV 24701
Trials and Tribulations
one businessman’s micro blues
Smoky Mountain Aero. With about $200,000 per month in aircraft loans— not to mention gasoline sales, service, instruction, charter,
rentals, tie-down and advertising— SMA uses a lot of business computer power.
Y ou see, I never had any in-
tention of using my micro
in my business; I just wanted it
to play with,” Jim Sexton tells
me. I was surprised because
other people had said Sexton’s
Maryville, Tennessee, flying
service uses a sophisticated
microcomputer business
system.
‘‘When I found out you could
do some useful things with it,”
he continues, ‘‘I moved the unit
from home out to the airport. I
kept expanding it, and the next
thing I knew I couldn’t take it
back home anymore.”
Jim’s background in elec-
tronics made him a natural
microcomputer user. He was a
radio-station engineer and
owned an avionics service
facility— among other accom-
plishments— before taking over
Smoky Mountain Aero, one of
the largest aircraft sales-and-
service operations in eastern
Tennessee. He still maintains
his avionics shop and offers
flight instruction and charter
service besides.
Smoky Mountain Aero’s
primary micro system is built
around an Imsai mainframe
with Seals memory, a PROM
board, Mits disk drive, a Lear
Siegler terminal and an LA-36
DEC Writer. Smoky Mountain
still shares a larger computer
with a Maryville bank, but the
micro handles mailing lists and
letter writing, aircraft tie-down
and maintenance records, air-
craft insurance, advertising
and sales duties. A second Mits
system is in the works.
Choosing hardware wasn’t
easy; writing software is a con-
tinuing, frustrating job; keeping
the machines on line is a big
headache. Even so, the com-
puter system is performing
useful tasks for the business.
“We’ve got enough stuff on it
so we couldn’t do without it
now,” Jim smiles wryly.
But, in spite of the smile, the
Smoky Mountain Aero story is
depressing at times, enough to
make a potential microcom-
puter user fold up his memory
cards and quit. On the other
hand, the Smoky Mountain folk
have a far-reaching view of the
micro business. With this tone
of ambivalence in mind let’s
start at the beginning.
Software Problems
Never one to do things in a
small way, Jim Sexton visited
various manufacturers and
retailers around the country,
flying to all the major cities in
the East and Midwest for equip-
ment demonstrations. The
results were less than satisfy-
ing. He settled on an early
Altair, then because of power
36
supply limitations, bought an
Imsai. Hungry for information
to turn his new toy into a useful
business tool, he continued
visiting suppliers.
Due partly to a lack of soft-
ware and apparent equipment
problems, Jim began buying
more hardware— a Processor
Technology SOL, lots of
memory, disks, etc.— but final-
ly settled on the Imsai/Mits
system he’s currently using.
The big problem was, and is,
getting useful software to run
on this system.
“I haven’t met anyone in my
travels— with the exception of
Altair Software in Atlan-
ta— who knows what they’re
doing,” Jim recalls. Even Altair
Software Distribution has its
problems, though. In Jim’s six
visits to buy a word-processing
package, the Altair Software
people were unable to show
him one that worked without
file-link errors or disk-drive
problems. “Every time they’d
try to demonstrate it, the thing
would switch off and come up
with some kind of error. They’ve
never been able to demonstrate
a working package to me,” he
says.
Jim tried several books of
programs, advertised to fill a
plethora of business and hobby
needs. He and his secretaries
spent hours keying in the
printed programs, only to dis-
cover that none worked proper-
ly. Based on a close analysis of
two of the programs— Depreci-
ation and Celestial Naviga-
tion— Jim believes the fault is
with the printed program. “I can
show you what is wrong with
the programs,” he says. “The
formulas are wrong. I don’t
know where they got those for-
mulas, but any basic finance
book can show you the error.”
Plus Ultra the Hobby Level?
Jim describes his experience
with micros so far as “discour-
aging,” but he hasn’t given up
on the idea of further utilizing
his system to make his busi-
ness more efficient. He
believes, however, there will
have to be some changes in
hardware and software before
personal computing will be
more than a novelty appealing
only to inveterate experiment-
ers and hobbyists.
Jim has a strong electronics
background, remember, but
each time a glitch develops he
has to go to Mits or Imsai for
help. Engineers and program-
mers naturally want to know
which memory locations he’s
using, how his software is con-
figured, etc. . . . questions that
only increase his frustration
level.
“When I bought the com-
puter I didn’t take time to learn
about memory locations and
octal and hexadecimal and all
this stuff; so they’d have to tell
me on the telephone which
switches to push, what to hit
next. Then I’d read the panel
and tell them which lights were
lit, and they’d tell me what they
thought was wrong.”
Some recent hardware ad-
vances— cheaper memory,
ROM programming, better disk
systems— perhaps have eased
some of the problems Jim suf-
fered through in the beginning.
Systems like the Commodore
PET and the Radio Shack
TRS-80 are welcome entries to
the computer field, but Jim Sex-
ton still believes the micro in-
dustry is too hardware oriented
and suffers from a hobbyist
mentality.
“The problem has been,” he
says, “that the micros have
been designed for the ex-
perimenter— where cost is a
major factor. I don’t think it
matters much what it costs,
let’s get one that’ll do the
job— make it reliable and easy
to operate— and see if busi-
nesses can afford it.”
The business market is
where Sexton envisions the
future of the micro industry. His
business activities over the
years have been varied: grocery
stores, filling stations and
other small operations that ran
concurrently with other oc-
cupations. He’d buy a business
that was in trouble, straighten
it out, then sell it for a profit.
This kind of transformation, he
says, is relatively easy because
the same problem usually ex-
ists: a lack of knowledge of
what it is costing to do busi-
Jim spends at least $500 per month on outside computer services
from a local bank. He gets printouts like this every two weeks to
help keep track of aircraft loans, business profits and tax infor-
mation. The printouts also list aircraft operating records, main-
tenance information and pilot time. He'd like to use his own
microcomputer for daily information, but he can’t trust it.
ness. ‘'Sometimes it is very
hard to know what each sale
actually costs.”
Jim believes a computer sys-
tem that worked, coupled with
a reliable software package
could encourage more people
to get into small business and
be successful. “I see the com-
puter as a way of keeping them
from failing,” he says, but
present systems would only
“heighten the frustration
level.”
Where It Came From;
Where It’s Going
What initially sold Jim on
microcomputers was a lot of
potential at a reasonable price.
He had checked out large com-
puter systems from IBM, Bur-
roughs and others, and the low-
est-priced package he could
get cost about $58,000. The
basic hardware for that system
was only $12K, but enough soft-
ware and mass storage to do
the job for Smoky Mountain
Aero pushed the price up in a
hurry.
“Over the years, the big com-
panies, by withholding technol-
ogy, have been able to sell a
commodity at a very, very high
price,” Jim laments. “But what
they’re doing with their large
systems isn’t really out of
reach of the micros.” There’s
that ubiquitous software prob-
lem, though: so far, workable
software that will run reliably
on a micro has eluded Smoky
Mountain Aero.
What is the seemingly elu-
sive job Jim so desperately
wants his micro system to do?
It already is doing a great deal.
In addition to typing original
advertising letters and han-
dling maintenance, tie-down
and insurance records, the
Smoky Mountain Aero com-
puter keeps track of flight-train-
ing records for Veterans
Administration-supported stu-
dents. The VA requires a com-
plicated series of records on
37
each student to be filed each
month. At Smoky Mountain
there are at least 50 VA-certi-
fied students. “The computer
reduces what used to take
three days each month to a cou-
ple of hours,” Jim says. “Mine
has paid for itself in that alone
if I could do nothing else with
it.” Still, Jim would like to see
his system do more.
He’d like to eliminate use of
the bank’s computer entirely
and switch to in-house process-
ing for a sophisticated ac-
counts-receivable and costing
program for Smoky Mountain
Aero.
“We have an accounts-re-
ceivable program written for
use on our computer, but I’m
afraid to start using it. If we get
into any of these screw-
ups— dropping bits and such—
it could really ruin our account-
ing program, so we’re still go-
ing with the bank because I
trust the bank.”
So, for now, instead of
switching to an all-micro
system, Smoky Mountain Aero
is investing even more heavily
in a time-share system. The
micro will continue to handle
the duties it already performs
well: duties that wouldn’t mess
up the company’s entire book-
keeping system should some-
thing go awry. They’ll purchase
several video terminals to ac-
cess the bank’s computer
directly, perhaps using the
micro system for some internal
processing; that data could be
fed to the large computer in a
block.
“What I really want to do is
get a daily profit/loss state-
ment, which in this business is
very, very difficult,” Jim says.
“In this industry, we are both-
ered by so many taxes and reg-
ulations that to employ some-
one at minimum wage to work
the front counter and make
decisions on what is taxable
and what is not is difficult.”
He’d like to use a terminal at
the front desk tied to a comput-
er programmed to make those
decisions, keep track of state,
federal and local taxes and
print out profits at the end of
each day. Currently, he gets
this kind of information only
every two weeks from the
bank’s computer. Smoky Moun-
tain Aero also has a program to
do this on the micro, “but,” Jim
reiterates, “I’m afraid to use it
because of the unreliability of
the computer.”
Jim Sexton has rejected the
idea of buying or leasing a big
machine. The cost is too high,
and he already has committed
around $8000 to micros.
Besides, he still believes a
microcomputer is capable of
the job he wants done at a
reasonable price. “I’m willing
to pay a bunch of money to
develop a computer system
that’ll work,” he admits.
And he’d like to see better
documentation with the equip-
ment already available. One of
the problems he had in the
beginning was with his Altair
2SI/0 boards. He couldn’t make
either of his two boards switch
to the second port; so when he
wanted to change from his
ADM-3 video terminal to his
DEC Writer, he’d turn off the
computer, pull out the I/O board
with the first port configured
for the ADM-3 and plug in one
set up for the printer. After a
year of that— even with
repeated calls to Mits for
help— he wrote a program to
overcome an apparent hard-
ware problem.
“They kept telling me to read
the instructions. Well, the in-
structions say if you’re running
the thing one way to put a cer-
tain switch up, but to change
that for other conditions,” Jim
says. “They say, ‘Where is the
BASIC addressed?’ and I have
no idea. I say, ‘How the hell do
you tell?’”
With the kind of software Jim
wants to run, an important
capability is to read informa-
tion off a disk or tape, update
the information and put it back
in the same place without
destroying what is on either
side. Documentation with his
disk system is so unclear that
so far he has been unable to
make it work that way.
“They use terms I simply
don’t understand. If they’d
charge me, say, $1800 for the
disk system, then another
thousand for a six-page booklet
of operating instructions I
could understand, then I’d be
willing to pay the extra thou-
sand bucks.”
Micros Should Get
Down to Business
Well, I said in the beginning
this story would be frustrating,
discouraging, confusing. It is
doubly so because I’ve talked
with potential micro users and
owners who say the same
thing: they simply can’t under-
stand the instruction manuals.
The manuals are either poorly
written and illustrated, or writ-
ten in such technical terms that
only a designer or programmer
with considerable experience
could understand them.
“It seems the micro industry
is just not conducive to
business applications,” Jim
observes. “You really have to
want to fool around with a com-
puter, spend hours to make it
work; I don’t think businesses
Jim Sexton has spent hundreds of hours trying to develop hardware and software for his business.
His experience with micros, however, has been discouraging. “I’d be better off with a mini. That’s
what I’d recommend to anybody interested in their own business computer,” he says. Even so, the
investment has been a good one. The computer has paid for itself: “ I’m not disappointed in my in-
vestment, it has been worth it. It just falls way short of what I’d like it to do.”
38
Selling and servicing equipment such as this $475,000 Cessna 421
is only one facet of the Smoky Mountain Aero operation. Yet a
single sale can involve complicated loan, trade-in and tax records,
which can be made manageable only with a computer.
are willing to accept that.
A few businesses— Smoky
Mountain Aero for one— have
accepted it, but only because
the owners have an interest in
computers. Most businesses
are interested only in what the
computer can do for them, not
how many bytes it stores, what
chip it uses or how big the
power supply is.
The American public is in-
creasingly aware of computers’
power, and businesses are ex-
pecting more and more from
these machines they’ve heard
so much about. The heavy-
weights in the computer in-
dustry aren’t serving the small
businessman. Micro hardware,
on the other hand, is available
to serve a wide mixture of busi-
ness and personal needs at
reasonable prices. Workable,
versatile, affordable software
to serve a variety of small
operations hasn’t yet arrived.
“The big-computer industry
goes for a large General
Motors-type company,’’ says
Jim. “They develop the soft-
ware needs of that company,
and if somebody else can fit it,
fine. If they can’t, it’s just too
bad. I’d hate to see the micro in-
dustry get to that point.”
Jim Sexton’s working on the
problem. He has hired a full-
time programmer in an effort to
put his $8K worth of hardware
to full use. Right now they’re
working on a parts-inventory
program and trying to polish
other aircraft-industry pro-
grams so they will run reliably
on a variety of microcomputers.
Already Jim has what he calls a
“pretty good software collec-
tion” he hopes will benefit
other Fixed Base Operators
(FBO).
“There are 500 to 600 FBOs in
this country, and maybe 50 of
them would be interested in
some of the things I have.
That’s not many units, but I’m
trying to help aviation more
than I am the computer in-
dustry,” he says.
Small-computer stores are
doing a booming business.
They’re selling the hardware
faster than they can get it from
manufacturers. But retailers
and manufacturers might do
well to note the experiences of
businesses like Smoky Moun-
tain Aero and take a dedicated
interest in turning around what
could be a disastrous trend. If
software development doesn’t
keep up with technical develop-
ment, we may be in for a user
backlash that could set the in-
dustry on its ear. Jim Sexton is
not the only businessman with
mixed sentiments toward the
microcomputer phenomenon. ■
MICROPROCESSORS
FROM CHIPS TO SYSTEMS
Rodnay Zaks
420pp, 150 illustr, ref C201 $9.95
Our Bestseller. This book is the result of the
author’s experience in teaching microprocessors
to more than 2000 persons. It presents a com-
prehensive introduction to all the aspects of
microprocessors, from the components to the
assembly of a system. The difficulty of each
chapter is graduated from the basic concepts
to the actual technical details. It is read by
students, technicians, managers, engineers,
educators, doctors and by all those who wish to
understand rapidly and efficiently all the important aspects
of microprocessor use, selection, or application.
CONTENTS: Fundamental Concepts... Internal Operation of a
Microprocessor... System Components... Comparative Micro-
processor Evaluation... System Interconnect... Microprocessor
Applications... Interfacing Techniques... Microprocessor Pro-
gramming... System Development... The Future...
ALSO AVAILABLE: MICROPROCESSOR ENCYCLOPEDIA
• VOL I: 8-BITS, ref E8 $15.00
• VOL II: BIT-SLICE, ref E5 $15.00
TO ORDER
• BY PHONE: call (415) 848-8233
BankAmericard/Mastercharge accepted
• SHIPPING: no charge when payment
included.
ADD: $1.00/book for fast shipping.
• TAX: in California, add sales tax.
• OVERSEAS: 66 French Francs. Write to:
SYBEX-EUROPE, 313 rue Lecourbe
75015 - PARIS, France Tel:(1)8282502
MICROPROCESSOR
INTERFACING
TECHNIQUES
Austin Lesea, Rodnay Zaks
350pp, 320 illustr, ref C207 $9.95
1
microprocessor
interfacing
techniques
FROM KEYBOARD TO FLOPPY DISK: All the
basic concepts and techniques for assembling
a complete microcomputer system are presented
in detail, with over 320 illustrations or diagrams,
including teletype, keyboard, floppy disk, CRT
display, analog / digital interfacing. One chapter
is dedicated to the bus standards, including
S100, and IEEE488, and one chapter to testing
and debugging.
CONTENTS: Introduction... Assembling the Central Processing
Unit... Basic Input-Output... Interfacing the Peripherals... Analog
Conversion... Bus Standards... Case-study: a 32-channel Multi-
plexer. . . Digital Trouble-Shooting. . . Conclusion-Evolution.
DISTRIBUTION /TRANSLATION INQUIRIES INVITED
2161 Shattuck Ave.
Berkeley, Ca 94704
Tel: (415) 848-8233
NAME POSITION
COMPANY
ADDRESS
CITY
□ Send me:
□ Payment enclosed □ Bill company □ Bill me
ADD $1.00 / BOOK FOR FAST SHIPPING
□ charge my □ Visa □ Master charge
□ Number Exp date
Signature _ □ Send catalog. S42
STATE/ZIP .
□C201 DC207 □ Other: .
39
THE TOTAL
COMPUTER SYSTEM
With RCC's REX you don't need to buy CPU boards and memory boards and parallel I/O boards and video display boards, and motherboards
and serial I/O boards and disk interface boards and cabinets and chassis and keyboards and stuff!
REX is the complete small computer, designed for the user who wants to buy a system with enough power to run FORTRAN ... or Disk
BASIC ... or Z80 Assembler .
The REX system costs $2495. Sure that's more than a kit, but look what you get!
Z80A CPU (4 MHz)
24 K of RAM
S-100 Bus
Microfloppy drive
Bootstrap monitor
B'standard
©standard/ expandable to 64K right on the
motherboard
©standard/ up to 5 slots available in basic
cabinet. . .expansion cabinet
available soon
&one standard, one optional/ record on
both sides of the diskette; 75K
bytes per surface
53-standard/ in a 2708 PROM. No switches;
no buttons
©standard
Powerfail, 60 Hz
timebase &
vectored interrupts
Black & white video B'standard/ 128 character set, 24 line,
interface 80 char/line, full software
video control
DMA channels B'standard/ control lines on the bus
Microfloppy
interface
ASR-33 type
keyboard
Walnut side-panel
cabinet
15 Amp power
supply
B'standard/ includes 8 bit parallel I/O port
B'standard/ optional typewriter keyboard
with numeric 10 key pad
B'standard
©standard
Your REX system is shipped assembled, tested and burned-in.
You just add a video monitor (or an RF modulator and your
home TV) and select one of our optional software packages
($195) and you are computing!
Accounting programs, analysis, games ... no problem! Your
basic REX system can handle them. Of course, we offer expansion
capabilities too. Options like:
□ Second microfloppy drive $450
□ Programmable CRT character generator $ 45
□ 12 inch CRT monitor (green phosphor, matching REX
wood-side cabinet) $212
□ 8K x 8 RAM expansion $145
□ 16K x 8 RAM expansion $285
□ Serial I/O channel (installed on initial order) $ 35
□ RF modulator (for connecting REX to home TV) ... $ 35
□ Dot matrix printer (120 cps>80/96 column) $850
□ Expansion cabinet for printer and two large
floppy drives $175
The software is super! Take your choice for $195!
□ ANSI FORTRAN IV, Disk Operating System, File Management
System, Linking Loader, Text Editor, Utilities & Diagnostics
□ Extended Disk BASIC, Disc Operating System, Text Editor &
Diagnostics
□ Z80 Macro Assembler, Disk Operating System, Text Editor &
Diagnostics
See the REX system today at your local computer store. Or call
RCC toll free at (800) 553-1863.
TM
realistic controls corporation
404 West 35th Street, Davenport, I A 52806
Toll Free (800) 553-1863 In Iowa (319) 386-4400
See the Rex
at these dealers
ARIZONA
Tempe Computerworld, Inc.
( 602 ) 944-6070
Tempe Phoenix Group. Inc.
( 602 ) 894-1129
CALIFORNIA
Escondido Business Enhancement Compuservice
( 714 ) 741-6335
Huntington Beach Cybernetics, Inc.
( 714 ) 848 1922
Lawndale Jade Computer Products
( 213 ) 679-3313
San Diego Byte Shop
( 714 ) 565-8008
San Francisco Computer Store of San Francisco
( 415 ) 431-0604
San Jose Byte Shop III
( 408 ) 377-4685
San Rafael Byte Shop-Computer Store
( 415 ) 457-9311
Santa Clara Byte Shop-Computer Store
( 408 ) 2494221
Sherman Oaks Peoples Computer Shop
( 213 ) 789-7514
Tarzana Tech-Mart
( 213 ) 344-0153
COLORADO
Englewood Byte Shop, Inc.
( 303 ) 761 -6232
IOWA
Davenport Computer Store of Davenport
( 319 ) 386-3330
KANSAS/MISSOURI
Overland Park Personal Computer Center, Inc.
( 913 ) 649-5942
MARYLAND
Baltimore Computer Workshop of Baltimore, Inc.
( 301 ) 486-5350
Rockville Computer Workshop, Inc.
( 301 ) 468-0455
MICHIGAN
Royal Oak Computer Mart of Royal Oak
( 313 ) 576-0900
MINNESOTA
Minneapolis Computer Depot
( 612 ) 927-5601
MISSOURI
Parkville Computer Workshop of Kansas City, Inc.
( 816 ) 741-5055
NEBRASKA
Lincoln Microtech, Inc.
( 402 ) 423-8404
Omaha Omaha Computer Store
( 402 ) 592-3590
NEW YORK
New York Computer Mart of New York
( 212 ) 686-7923
OHIO
Lyons Affordable Computers
( 419 ) 923-2821
OREGON
Beaverton Byte Shop
( 503 ) 644-2686
Portland Byte Shop
( 503 ) 223-3496
PENNSYLVANIA
Pittsburgh Computer Workshop of Pittsburgh, Inc.
( 412 ) 327-0455
TEXAS
Dallas Computer Shops, Inc.
( 214 ) 234-3412
Dallas Rondure Company
( 214 ) 630-4621
Houston Interactive Computers
( 713 ) 772-5257
San Antonio The Computer Shop
( 512 ) 828-0553
San Antonio Computer Solutions
( 512 ) 828-1455
VIRGINIA
Springfield Computer Workshop of Northern Virginia, Inc.
( 703 ) 321-9047
WASHINGTON
Bellevue Byte Shop
( 206 ) 746-0651
Bellevue The Microcomputer Center
( 206 ) 455-3710
WISCONSIN
Janesville Austin Computers
( 608 ) 752-6002
' realistic controls corporation
Look To The Horizon.
The North Star
HORIZON Computer.
HORIZON ™
To begin programming in extended BASIC, merely add
a CRT or hard-copy terminal. HORIZON- 1 includes a
Z80A processor, 16K RAM, minifloppy tm disk and 12-
slot S-100 motherboard with serial terminal interface— all
standard equipment.
And, Look To Computer Enterprises
For The Lowest Prices & Fastest
Delivery On North Star Horizon
Computer System:
Credit
Cash
Card
Dtocount
Price
Price
Horizon 1 System (one drive) kit .
. . $1497
$1439
Horizon 1 System (one drive)
assembled
. . $1777
$1709
Horizon 2 System (two drives) kit
. . S 1 87 1
$1799
Horizon 2 System (two drives)
assembled
..$2199
$2114
ALSO: Run TDL Software On Your
Horizon...
TDL Package A with Super Bask . . $ 228 $ 219
For Your Serial Terminal:
Lear Siegler ADM-3A (kit) $ 728 $700
Lear Siegler ADM-3A (assembled) .$ 858 $ 825
Call Or Write Today
For Computer Enterprises'
FREE Spring Computerlogue!
IMSAI PCS-80/30 kit
.$1097
$1055
IMSAI PCS-80/15 kit
$
748
$
719
IMSAI VIO-C kit
S
297
$
286
TDL ZPU Board (assembled)
$
183
$
176
TDL XITAN ALPHA 1.5 kit
$
812
$
781
Seals 8k 250ns RAM (assembled) . .
$
203
$
195
Dynabyte 16k 250ns Static RAM
(assembled)
S
577
$
555
Cromemco Z2D kit
,$1399
$1345
Cromemco Bytesaver less Proms kit
$
136
$
131
Shipping charges: $10 per CPU on larger units; SI. 50 per kit. $2.00 min.
per order.
Delivery is stock to 30 days on most items. Shipment is immediate for
payment by cashier’s check, money order or charge card. Allow 3 weeks
for personal checks to clear. N.Y. State residents add approp. sales tax.
Availability, prices and specs may change without notice.
computer
enterpriser
Operating Hours:
M-W 10-5 E.S.T.
TM TH-F 10-9 E.S.T.
Closed Sat. & Sun.
P. O. Box 71 Fayetteville, N.Y. 1 3066
Phone (315) 637-6208 Today!
Robert Baker
15 Windsor Drive
Atco NJ 08004
Writing Diagnostic
Routines
while your machine is running
E ver since personal com-
puter systems first
caught on, the area of diag-
nostic software has been over-
looked. Few companies
producing kits and hobby
systems offer diagnostic pro-
grams to test and debug their
products should a problem be
suspected. Possibly such pro-
grams haven't been offered
yet because of the degree of
customizing each user per-
forms while assembling his or
her system. Because of
competition in the market,
most systems are a conglom-
eration of bits and pieces
interconnected for a partic-
ular application. Thus, most
hobbyists are forced to write
their own diagnostic or
maintenance programs (usual-
ly after a major problem
develops) without really
knowing what they're doing.
It's very difficult to debug
software if the hardware is
not working properly!
On mini and larger size
computers, the manufacturer
usually provides various
diagnostic programs to be run
at regular intervals by the
user as a form of preventive
maintenance. The programs
are written to detect minor
faults before they degrade
system operation, and to help
isolate and debug major
problems when they occur.
Diagnostic Methods
One of two approaches is
generally taken for writing
and using diagnostic pro-
grams. A bottom up approach
starts by testing the smallest
entity in the system then
using that proven-good device
to test the next device in the
system until all devices have
been tested. Individual com-
ponents are then tested in
clusters or subsystems, and
finally the entire system is
tested, or exercised, as a
whole.
A top down approach, on
the other hand, starts by
running a system exerciser to
test all devices at once and
isolate a problem to a given
subsystem. More detailed,
device-dependent programs
are then run for the particular
faulty device or subsystem to
further isolate and help debug
the problem. Once the prob-
lem has been corrected and
the device-dependent tests are
passed, the system exerciser
can be run again to verify
that that was the only system
fault.
A bottom up approach
requires the least amount of
working hardware to be use-
ful, but a top down approach
takes less time to isolate a
given system fault; so there
are trade-offs. For either
approach, the actual pro-
grams could be similar, de-
pending on the system and
the application.
Writing A Test Program
Why not write a collection
of test programs while you
have a working system to try
them on? Debug your pro-
grams thoroughly when writ-
ing them so you're sure that
any problems detected are
caused by hardware and not
software. Try a bottom up
approach first as this should
make the programs easier to
write. Start out with a few
simple programs to check the
CPU machine instructions,
checking operands, condition
codes, etc. Then check data
paths to and from the CPU
and the various control logic,
trying different bit patterns
to check for shorted lines.
For convenience, you may
want to create some of these
programs in ROM and have
them permanently available.
Loading programs would
require a major portion of the
CPU to be in working order,
so using ROM would elimi-
nate that problem.
A quick memory check
can verify that RAM memory
is working correctly by writ-
ing all zeros and all ones to
each location, reading it back,
and comparing the data. You
may want to check another
memory pattern such as al-
ternating ones and zeros
(10101010) as well as check-
ing memory addressing logic
by insuring a test pattern had
not been written into another
location.
After the CPU and mem-
ory have been tested, you can
then proceed to test any
other devices you may have
in your particular system.
Test each device separately
and thoroughly before going
to the next. For starters, try a
42
program to test your CRT
display or video terminal
with:
• a character generator check,
full lines of each character.
• display memory test (swirl
pattern). The first line is a
full character set. Each line
after the first starts with the
next letter in the character
set after that used on the line
above it. Therefore, each
character will be on a diag-
onal, and will appear in each
storage location as the test is
run, and the display scrolls.
For example:
abcdefg
bcdefg
cdefg .
defg . .
Other patterns can be added
to test special features, etc.,
depending on the particular
display. To test a keyboard
you can try a program that:
• displays on your terminal
or CRT the code for the key
depressed and the actual char-
acter.
• asks you to type each char-
acter in a set sequence and
checks the code received.
Another useful test could
print continuous lines of any
character typed in by the
user. When another is typed
in, the printer would change
to that character. Similar
tests can be written to check
the particular features or
functions of other displays
and terminals. Use easy-to-
recognize patterns on printers
or terminals and keep the
tests simple!
If you have tape drives,
cassettes, floppy drives, opti-
cal readers or joysticks, don't
forget to test them also. Test
every device in your system
thoroughly, one at a time.
Later you can add a simple
exerciser to get everything
working at once and check
for device interactions. My
advice is to save this for later
when you start to get a feel
for what you really want to
accomplish and how you
want to control your test
programs.
Make It Useful
For whatever devices are
being tested, certain basic
features should be included
for your own convenience
and increased usefulness of
the programs. Each test
within a program should give
a clearly defined indication
when an error is detected.
This can be an error message
on a terminal or simply a
machine stop at a specific
address. If error stops are
used, separate halts or stops
should be used for each
possible error so the address
at which the machine stops
will indicate what error was
detected. For added con-
venience you may even want
to generate an error dic-
tionary to list and describe
each possible error halt and
give some possible causes or
cures. Also, keep a log of the
errors and causes detected by
your programs for later refer-
ence. They may save you
from debugging the same
problem again several months
or years later.
Another desirable feature
should be the capability to
loop on any test for scoping
purposes, possibly with a
sync pulse generated at the
start of each pass. Other
features can be added as
desired, depending on your
particular system and how
you want to test it. Try to
keep it simple but flexible,
and easy to use.
Plan ahead and prepare
yourself for the inevitable.
Sooner or later you're bound
to have a hardware problem,
and your local TV repairman
will probably not be able to
help you, let alone know
what you're talking about.
An even better idea: Run the
tests periodically and catch
small problems before they
become major ones. Your
time spent writing diagnostic
programs for your system
now will be repaid many
times in the future. ■
the Computer Store
the store for the professional
Offers
A Unique
Profit
Opportunity
FOR
Software
Companies
Systems Houses
Computer
Retailers
The computer Store’s Dealer Affiliate
Program features the Data General
microNOVA computer line and estab-
lished professional business applica-
tions software for resale to the explo-
sive small business computer market.
Sales and technical training included.
For details:
Call Paul Conover
617 - 272-8770
the Computer Store
120 Cambridge St.
Burlington MA 01803
617-272-8770
Experiments in Software
serial to parallel conversion
Dan Stogdi/I
182 Victoria Street
St. Marys Ontario
Canada NO M 2 VO
A s most hobbyists are
aware, microcom-
puters generally employ
parallel data to carry out
their internal transactions.
That is, the CPU expects all
bits of a given data word to
exist simultaneously for a
finite length of time, one bit
on each of the parallel wires
of the data bus. Nevertheless,
in order to communicate with
the outside world, this
parallel data must frequently
be converted to serial form in
which each bit of a given data
word is transferred sequen-
tially between the micro-
computer and its external
devices. Technology has
provided us with at least two
hardware devices (the UART
and the ACIA) designed to
achieve this end. Instead of
hardware, however, software
may also be employed to
attain the same results.
Although there have been
several articles on this subject
in the literature, most of
them contain a cautionary
note related to the param-
eters associated with the soft-
ware timing loops. By way of
illustration, the typical
routine which converts
incoming serial data to
parallel form does so by
testing the middle of each bit
in order to determine
whether it represents a binary
1 or 0. Following examina-
tion of the center of one bit,
the program will loop
through a time delay routine
for a fixed length of time.
This length of time corre-
sponds to the duration of one
bit. Upon exiting from the
loop, the program tests the
next bit, loops . . . and so on.
The requisite loops are
dependent on the speed with
which the CPU executes its
instructions (i.e., does it have
a 1 MHz clock, 2 MHz clock,
etc?) and by the access time
of the memory. With slower
memories it is general
practice to insert one or more
wait states before the mem-
ory is read. As a consequence,
to some extent serial-to-
parallel conversion routines
tend to be machine or system
specific.
In this article, I intend to
describe a simple serial-to-
parallel routine, READ, and a
Symbolic Address
Location
Machine Code
Mnemonic
Comments
CALIBRATE
000-200
041
LXI H
Initialize the storage
201
100
location for the
202
000
critical value
TST
203
333
IN
Input the parallel
204
005
port
205
346
ANI
Set up a
206
001
mask
207
302
JNZ
If no start bit
210
203
TST
go back to TST
211
000
212
001
LXI B
Initialize the B-C
213
000
register pair to
214
000
zero
215
315
CALL
Call subroutine
216
261
TIMER 1
217
000
220
161
MOV M,C
Store C register
221
043
INX H
Bump the pointer
222
160
MOV M,B
Store B register
223
166
HLT
Halt
...
Unused memory space
TIMER 1
000-261
003
INX B
Increment B-C pair
262
333
IN
Input the parallel
263
005
port
264
346
ANI
Set up a
265
001
mask
266
177
MOV A, A
Extra-parallels TIMER 2
267
267
ORA A
Extra-parallels TIMER 2
270
312
JZ
If no data bit
271
261
go back to TIMER 1
272
000
273
311
RET
Return to CALIBRATE
Table 1. This routine determines the critical value and stores it in addresses 100 and 101 (octal) on page 0.
44
paral lel-to-serial routine,
WRITE, which allows my
Altair 8800 (through one of
the parallel ports on a
Processor Technology 3P+S
interface) to converse with
my CRT terminal which
employs an RS232C serial
I/O. Of greater importance, I
will outline a simple tech-
nique by which anyone can
determine the values of the
loop counters for any partic-
ular system.
Some Preliminaries
Fig. 1 illustrates the simple
connections made to the
3P+S in order to boost the
TTL signals at the parallel I/O
port to RS232C level
capable of interfacing with
my terminal. As can be seen,
for outputting data, one
output line of a parallel I/O
port (available at J1 pin A) is
fed back to the input of an
unused gate of 1C 5, the 1488
which normally boosts the
3P+S's UART output to
RS232C level. For inputting
data, one input line of a
parallel port (available at J2
pin Z) is connected to the
Fig. 1. The simple modifications to the Processor Technology 3P+S
interface.
Symbolic Address
Location
Machine Code
Mnemonic
Comments
READ
000-000
305
PUSH B
Save these
001
325
PUSH D
registers
002
345
PUSH H
003
021
LXI D
Set up number
004
010
of word bits in reg. E
005
000
Clear reg. D
TESTR
006
333
IN
Input the
007
005
parallel port
010
346
ANI
Set up a
Oil
001
mask
012
302
JNZ
If no start bit
013
006
go back to TESTR
014
000
015
001
LXI B
Initialize B-C pair
VALU 1
016
224
with loop counter
017
002
value
020
315
CALL
Call TIMER 2 and
021
061
loop for awhile
022
000
NEXT
023
333
IN
Input the parallel
024
005
and fetch a data bit
025
346
ANI
Set up
026
001
mask
027
202
ADD D
Add D reg. to A reg.
030
017
RRC
Shift reg. A to right
031
127
MOV D,A
Save byte in reg. D
032
001
LXI B
Initialize B-C pair
VALU 2
033
270
with loop counter
034
001
value
035
315
CALL
Call TIMER 2
036
061
and loop
037
000
for awhile
040
035
DCR E
Decrement bit counter
041
302
JNZ
Fetched all bits?
042
023
No. Go back to next and get
043
000
another bit
044
172
MOV A,D
Put word in Acc.
045
247
ANA A
Set the flags
046
342
JPO
If parity is odd
047
057
jump to WRONG
050
000
051
346
ANI
Strip the parity
052
177
bit
053
341
POP H
Retrieve the
054
321
POP D
previously stored
055
301
POP B
stored registers
056
311
RET
Return to calling program
WRONG
057
166
HLT
Safety halt
060
000
NOP
Unused
TIMER 2
000-061
013
DCX B
Decrement loop counter
062
333
IN
Extra-parallels TIMER 1
063
005
Extra-parallels TIMER 1
064
346
ANI
Extra-parallels TIMER 1
065
001
Extra-parallels TIMER 1
066
171
MOV A,C
Fetch counter value
067
260
ORA B
Are B and C zero?
070
302
JNZ
No. keep looping
071
061
072
000
073
311
RET
Return to calling program
Table 2. This program converts incoming serial data
to parallel format and checks for even parity. (Note that on my system 1 obtained a
critical value of 270-001 f as shown by VALU 2.}
45
2 stop
bits
parity
bit
bit 7
bit 6 bit 5 bit 4 bit 3 bit 2
1 i 1 i
bit 1
start
bit
Resting Level
Logic 1
Logic 0
9.09 ms
1 00 ms
Fig. 2. The bit composition of the ASCII encoded character A (capital ), and its associated timing values when transmitted at a rate of 1 10 baud.
Note that the resting level is always logic 1, while the start bit is always logic 0.
output of an unused gate of
1C 10, the 1489 which
converts the incoming
RS232C signal to TTL level
compatible with the UART's
input.
Owners of other parallel
I/O boards and/or current
loop type terminals can also
interface them readily. Back
issues of 73 and Kilobaud
contain a number of circuits
based on the hardware
UART. Simply borrow that
part of the circuit which is
responsible for changing the
UART's I/O signals to the
appropriate drive level con-
sistent with your particular
peripheral.
The Critical Value
Fig. 2 depicts the bit com-
position of the ASCII char-
acter A (capital), and its
timing values when it is trans-
mitted serially at a rate of
110 baud. Note that each bit
occupies a time of 9.09 ms
and that the resting level is
always logic 1, while the start
bit is logic 0. In addition,
observe that the first data bit
of the A is always a logic 1
. . . this is very important in
determining what I call the
critical value.
Table 1 contains a short
program labeled CALI-
BRATE. This program
operates by testing for the
start of transmission of the
character A (for example)
when it is transmitted by the
terminal. Upon detecting the
logic 0 start bit (see Fig. 1), it
calls up a subroutine TIMER
1, which alternately incre-
Symbolic Address
Location
Machine Code
Mnemonic
Comments
WRITE
000-100
305
PUSH B
Save these
101
325
PUSH D
registers
102
345
PUSH H
103
247
ANA A
Set the flags
104
352
JPE
If even parity
105
111
in Acc. then
106
000
skip to EVEN
107
356
XRI
Make Acc.
110
200
even parity
EVEN
111
127
MOV D,A
Save the word in reg. D
112
036
MVI E
Set up number of
113
010
word bits in reg. E
114
257
XRA A
Zero the Acc.
115
323
OUT
Out this as a start
116
005
bit to the terminal
117
001
LXI B
Initialize the
VALU 2
120
270
loop counter
121
001
122
315
CALL
Call TIMER 2
123
061
and loop
124
000
OUTIT
125
172
MOV A,D
Move the word to Acc.
126
323
OUT
Output a character
127
005
to the terminal
130
017
RRC
Shift reg. A to right
131
127
MOV D,A
Save the word in reg. D
132
001
LXI B
Initialize the
133
270
loop counter
134
001
135
315
CALL
Call TIMER 2
136
061
and loop
137
000
140
035
DCR E
Decrement bit counter
141
302
JNZ
Outed all bits?
142
125
No. Go back to outit and
143
000
keep outputting bits
144
076
MVI A
Put a 1 in Acc.
145
001
Use as a stop bit
146
323
OUT
Output the stop bits
147
005
to the terminal
150
001
LXI B
Initialize the
VALU 3
151
160
loop counter
152
003
for two stop bit lengths
153
315
CALL
Call TIMER 2
154
061
and loop
155
000
156
341
POP H
Retrieve the previously
157
321
POP D
pushed registers
160
301
POP B
161
170
MOV A,B
Restore the Acc.
162
311
RET
Return to the main program
Table 3. This program converts parallel data to serial format and outputs it to the terminal.
46
merits the value in the B-C
register pair and tests for the
beginning of the first data bit,
a logic 1 (see Fig. 1). Upon
detecting the logic 1 start bit,
TIMER 1 is exited with a
relative value of the bit length
in the B-C register pair.
Should you run the program
several times, you will see
that the critical value stored in
addresses 100 and 101 (octal)
will remain quite stable.
(Note that address 100 con-
tains the least significant bits
and that address 101 contains
the most significant.) If you
have access to a terminal with
adjustable baud rates, you
will notice that the critical
value becomes progressively
smaller as the baud rate
increases.
The READ Program
Table 2 contains the pro-
gram READ which converts
incoming serial data to
parallel format. It assumes
that the parallel I/O is at
address 005 and, further-
more, that the serial data is
entering through line zero of
the input port (DIO). VALU
2 is the previously obtained
critical value. VALU 1 repre-
sents the critical value con-
verted to decimal, multiplied
by 1.5, and converted back to
octal.
TIMER 2 may appear to
contain some irrelevant
instructions. Not really.
TIMER 1 and TIMER 2 were
fashioned in such a manner
that the critical value
obtained with TIMER 1
would be directly applicable
with TIMER 2; their respec-
tive instructions were juggled
around while insuring timing
compatibility between the
two routines.
The READ program tests
for the beginning of the start
bit and then circulates
through the loop for a time
equal to 1.5 bit times
(governed by VALU 1). At
this point, the middle of the
first data bit is tested. Sub-
sequently, TIMER 2 loops for
1 bit length (governed by
VALU 2, the critical value),
exiting near the center (hope-
fully) of each successive bit
until all bits have been tested
and the word assembled in
parallel.
The WRITE Program
Table 3 contains the pro-
gram WRITE, which converts
a parallel data word to serial
format and outputs it to the
terminal. It also assumes that
the I/O port is located at
address 005 and that serial
transmission to the terminal
is occurring through line zero
of the parallel outport (DOO).
VALU 3 produces the stop
bits and is equal to the
critical value converted to
decimal, multiplied by 2, and
converted back to octal.
Conclusions
Both routines assume that
they will be called up by
other programs to provide
I/O services; as such, the
states of all registers except A
and PSW are saved on the
stack so as to not interfere
with other operations. As
provided, both routines
return to the main calling
program with the data word
in the accumulator. In addi-
tion, WRITE assumes that the
data word is in registers A
and B on being called. Lastly
the routines provide for
parity generation and check-
ing (even parity, in the
present case).
While described from a
viewpoint of being used with
a terminal, the routines
should lend themselves for
use as cassette I/O routines in
concert with appropriate
encoding/decoding schemes.
The software connoisseurs
among us could make them
more efficient memory-wise.
In addition, I am sure that
CALIBRATE and READ/
WRITE could be combined in
such a way as to make the
whole process automatic in
nature.
What initially started out
as an experimental project is
now in everyday use on my
system. By all means experi-
ment with them and have
fun. As ever, if you have any
questions drop me a line and
I will try to help out. Keep
on computing. ■
Peter Stark
Box 209
Mt. Kisco NY 10549
Computer Math Primer
beginner's introduction
to number systems
O nce you understand a
little about hex # that
will make computers a little
less magical and bring them
down to earth. So here goes.
Hex is short for hexa-
decimal, one of four number
systems used with computers.
The other three are binary,
octal and decimal. Each is
based on a particular number:
binary uses the base 2, octal
the base 8, decimal the base
10 and hexadecimal the base
16. Table 1 shows their rela-
tionship.
Notice that binary num-
bers only use 0 or 1 ; there are
no digits larger than 1. So,
the next number after 0 and
1 is 10 (0010), since we have
to skip over 2, 3 and so on.
(For most purposes, 10 and
0010 are the same; the be-
ginning zeros don't change
things.) After 10, binary
numbers skip to 1 1, then skip
to 100 and 101, etc. Since
101 is the fifth number after
0, it stands for 5.
There are no 8s or 9s in
octal numbers, so after 7
comes 10; after 17, 20; after
77, 100.
Decimal numbers go from
0 through 9. That's only ten
different digits, so after 9 we
have to start doubling up;
after 9 comes 10, and so on.
The hexadecimal number
system has 16 different digits.
The first ten are the same as
the ten digits of the decimal
system — plus there are six
more.
I think whoever devised
hex made a big mistake here.
Having six new digits, he
should have invented six new
symbols for them. Instead, he
simply gave his six new digits
old symbols: A, B, C, D, E
and F. Consequently, we have
a number system that goes
from 0 to F. (After F, by the
way, you have to start
doubling up as in decimal.
After F comes 10, which
corresponds to 16 in dec-
imal.)
Computers use binary
numbers for their internal
operations. Octal and hex
numbers are used by the
people who use computers
(not computers themselves)
because they are easier to see
and interpret (A3, for ex-
ample, is easier to read than
10100011). Computer people
use octal or hex numbers
rather than decimal because
their conversion to and from
binary is easy and fast. (Use
of these number systems
constitutes external oper-
ations.)
Whether octal or hex is
used depends partially on the
length of the binary numbers
used, and partially on per-
sonal preference. Computers
handle binary numbers of a
fixed length, called a word
length. Most hobby com-
puters use a word length of 8,
meaning they handle binary
numbers in groups of eight
digits. If the binary word
length is divisible by three,
octal is generally used; if it is
divisible by four, hex is
employed. If it is divisible by
both three and four, or by
neither, then it's a matter of
choice.
So, most hobby computers
with a word length of 8 use
hex (externally), although
there are exceptions: the
8008 CPU uses octal, as does
Heathkit's new H8. The
12-binary-digit (12-bit — a bit
is a binary digit) Intersil uses
octal.
Most hex or octal numbers
used with small computers
are small two-digit hex or
three-digit octal numbers.
Let's use those as examples of
how to do magic with them.
(Assume only positive inte-
gers for a starter.)
Converting Hex to Binary
For each hex digit, replace
it by its four-bit binary
equivalent from the table.
For example, to convert hex
A3, replace A by 1010 and 3
by 001 1 for a binary result of
10100011.
Octal to Binary
For each octal digit, re-
place it by the rightmost
three bits of the corre-
sponding binary number in
the table. To convert octal
243, for example, replace the
2 by 010, the 4 by 100 and
the 3 by 011. The complete
binary number is 01010001 1.
Now, for a second trick: This
procedure gives you a total of
nine bits, whereas most small
computers need only eight.
Fortunately, the leftmost bit
will usually be a zero, and so
?
48
Binary
Octal
Decimal
Hex
0000
0
0
0
0001
1
1
1
0010
2
2
2
0011
3
3
3
0100
4
4
4
0101
5
5
5
0110
6
6
6
0111
7
7
7
1000
10
8
8
1001
11
9
9
1010
12
10
A
1011
13
11
B
1100
14
12
C
1101
15
13
D
1110
16
14
E
1111
17
15
F
Table 1.
can be crossed out, giving the
final eight-bit answer of
10100011.
Binary to Hex
Starting from the right,
separate the binary digits into
groups of four; replace each
group by its hex equivalent.
Binary 01100100 would be
split into 0110 and 0100.
0110 is replaced by 6 and
0100 by 4, giving hex 64.
This conversion is easy if
the number of bits is any
multiple of 4. If not, you
must add zeros at the left
until it is. For instance, to
convert 11011 you first add
three zeros to make it
00011011, then split it up
into 0001 and 1011, and
finally convert to 1 B.
Binary to Octal
This is the same as the hex
conversion except that we use
groups of three bits. For
example, 01100100 is an
eight-bit number, and eight
bits cannot be separated into
groups of three; so we add an
extra zero to make it
001100100. We can then
break it up into 001-100-100.
We can convert each group
into octal using the table if
we note that 001 is the same
as 000 1 , or an octal 1 ; 1 00 is
the same as 0100 or an octal
4. Thus, binary 001100100 is
octal 144.
(Just a reminder: We are
working only with positive
whole numbers (integers)
now. Negative numbers or
fractions are a different ball
game.)
Binary to Decimal
The easiest way to do this
is to convert the binary
number to octal or hex first
and go to decimal from there.
If you insist on doing it di-
rectly, then here's how.
In a binary number, each
bit has a specific value; to
convert, you have to multiply
each bit by its value and add
the results. Starting from the
right, the values are 1, 2, 4, 8,
16 and so on — each value is
twice the preceding value.
To convert the binary
1011, transcribe the number
(spread it out), and under
each bit write its value.
10 11
8 4 2 1
Remember to start with a
value of 1 at the right. Now
multiply each bit by its value,
like this:
10 11
x8 x4 x2 xl
8 0 2 1
Finally, add 8+0+2+1 = 11.
(Check the table and you'll
see that 1011 in binary is 11
in decimal.)
To see how this works for
larger numbers, see Example
1. (With a little practice, you
can skip multiplying by zero.)
Add the results (128+
1 0
xl 28 x64
128
32+16+4) and you'll see that
binary 10110100 converts to
180 in decimal.
Octal to Decimal
This conversion is the
same as binary-to-decimal
except that the values of each
digit are 1, 8, 64, 512 and so
on - each value is eight times
more than the one before it.
To convert octal 264, for
example, use the preceding
technique
values.
with
the new
2
6
4
x64
x8
x1_
128
48
4
Now add up 128+48+4=180.
Hex to Decimal
Same as before, except the
digit values are now 1, 16,
256, 4096, etc. (remember to
start with 1 at the right) —
each value increases 16 times.
Hex 2C4 converts like this:
2 C( 1 2) 4
x256 x 16 xl
512 192 4
512+192+4=708.
Decimal to Binary
Convert to octal or hex
first; then convert the result
to binary. There is a direct
way, but it's likely to take
longer, and you will probably
make a mistake — so don't
bother.
Decimal to Octal
This conversion is done by
dividing the decimal number
by 8 and saving the remain-
ders. Keep doing this until
you get 0, and then put the
remainders together back-
wards.
This sounds crazy until
you see how it's done. Let's
say you want to convert 180
from decimal to octal. Start
by dividing 180 by 8. Don't
use your calculator for this
1
1
0
1
x32
x16
x8
x4
32
16
4
Example
1 .
because you will just get an
answer of 22.5, and then you
won't know what to do. Use
pencil and paper.
Eight goes into 180
twenty-two times, with a
remainder of 4. (After a while
you'll figure out how to use
your calculator for this, too.)
Put away the 4, and divide
the quotient by 8.
Eight goes into 22 twice,
with a remainder of 6. Save
the 6, and divide 2 by 8.
Eight goes into 2 zero
times, with a remainder of 2.
Since we are down to a
quotient of 0, we can stop
dividing by 8.
Now take the three re-
mainders (4, 6 and 2) and
write them backwards: 264.
This is your octal number.
Decimal to Hex
This conversion is the
same as decimal-to-octal,
except you divide by 16. For
example, to convert 180 to
hex, start by dividing 180 by
16.
Sixteen goes into 180
eleven times, with a re-
mainder of 4. Save the 4 and
repeat.
Sixteen goes into 11 zero
times, with a remainder of
11. Again, we stop dividing
when we get a quotient of 0.
Write the remainders down
backwards, but convert any
remainder above 9 to its hex
digit. In this case, the 11
converts to B, and the hex
answer is B4.
Complements
A complement is an
opposite; in the case of com-
puters, complements are used
for negative numbers. There
are two kinds of comple-
ments: one's complements
and two's complements. The
one's complement is easy to
find, but the two's comple-
ment is generally used.
0 0
x2 xl
49
We’ve gathered the family to show you why
PERCOM’s™ Number 1
in cassette data systems for microcomputers.
Pardon us for doing a little boasting, but we’re proud of our
family. Proud of each member’s reputation for performance and
reliability. And pleased that we can offer the best in cassette
data systems and data terminal interfacing at low, home-com-
puting prices.
It took more than guts and a little luck to
forge a position of leadership. We’re
number 1 because you get more when
you buy PERCOM™. The reason, simply,
is experience. Every product described
in this ad is based on nearly 10 years of
crucial involvement in the design and
manufacture of computer peripherals
that use cassettes for mass storage.
Experience. It’s why we developed a
more reliable data cassette for home
computing. Why our interfacing units
provide both cassette and data terminal
interfacing. Why you get the fastest,
most reliable cassette data rates from
PERCOM™. Experience. It’s the reason
for PERCOM™.
For your SS-50 bus computer — the
CIS-30+
For your data storage —
data cassettes
Pilon-30™
• Interface to data terminal and two cas-
sette recorders with a unit only 1/10
the size of SWTP’s AC-30.
• Select 30, 60, or 120 bytes per second
cassette interfacing, 300, 600 or 1200
baud data terminal interfacing.
• Optional mod kits make CIS-30+ work
with any microcomputer. (For MITS
680b, ask for Tech Memo TM-CIS-
30+ — 09.)
• KC-Standard/Bi-Phase-M (double fre-
quency) cassette data encoding. De-
pendable self-clocking operation.
• Ordinary functions may be accom-
plished with 6800 Mikbug™ monitor.
• Prices: Kit, $79.95; Assembled,
$99.95.
Prices include a comprehensive instruction
manual. Also available: Test Cassette, Re-
mote Control Kit (for program control of
recorders), 1C Socket Kit, MITS 680b mod
documentation, Universal Adaptor Kit
(converts CIS-30+ for use with any com-
puter). MIKBUG® Motorola, Inc.
For your S-1Q0 computer— the CI-812
• Both cassette and data terminal inter-
facing on one S-100 bus PC board.
• Interfaces two recorders. Record and
playback circuits are independent.
• Select 30, 60, 120, or 240 bytes per
second cassette interfacing, 110 to
9600 baud data terminal interfacing.
• KC-Standard/Bi-Phase-M (double fre-
quency) encoded cassette data. De-
pendable self-clocking operation.
• Optional firmware (2708 EPROM)
Operating System available.
• Prices: kit, $99.95; assembled,
$129.95.
Prices include a comprehensive instruction
manual. In addition to the EPROM Operating
System, a Test Cassette, Remote Control Kit
(for program control of recorders), and an 1C
Socket Kit are also available.
• Orders-of-magnitude improvement in
data integrity over ordinary audio cas-
settes.
• Pilon-coated pressure pad eliminates
lint-producing felt pad of standard
audio cassettes.
• Smooth pilon coating minimizes erra-
tic tape motion.
• Foam pad spring is energy absorbing.
Superior to leaf spring mounted pad
which tends to oscillate and cause flut-
ter.
• Five-screw case design virtually pre-
cludes deformation during assembly.
• Price: $2.49.
PERCOM™ products may be purchased
from home computer dealers nation-
wide, or may be ordered direct from the
factory.*
*Texas residents must include an
additional 5% for factory orders. MC &
Visa cards honored.
PERCOM™ ‘peripherals for personal computing'
PERBCM
PERCOM DATA COMPANY, INC.
DEPT. K
318 BARNES • GARLAND, TEXAS 75042
Phone:(214)276-1968
P7
50
Converting a binary num-
ber to one's complement.
Write down the binary num-
ber. Then invert each bit —
that is, change each 1 to a 0
and each 0 to a 1. For ex-
ample, the one's complement
of 10110 is 01001.
Converting an octal num-
ber to one's complement.
Write the octal number. Then
above each digit put a 7. Now
subtract each bottom digit
from the top digit.
To convert the octal 0145,
for example, you proceed like
this:
7
-0
7
-1
7
-4
7
-5
7
7
7
7
1
8
0
0
-0
-2
-3
-0
-7
-5
-0
-0
7
5
4
7
0
3
0
0
Example 2.
Remember that in the
binary number that's actually
in your computer, each 0 is
being inverted into a 1 as you
complement. Any extra zeros
you put in will produce extra
ones in the complement. For
example, octal 5 is binary
101. But it is also 0101,
00101, 000101, etc., since
putting extra zeros in front of
a binary number does not
change it. But look what
happens if you try to get the
one's complement (Fig. 1).
An octal 5 can have many
different complements; but
notice that the only differ-
ence between them is the
presence of extra ones at the
left. The solution is to use
only as many ones at the left
as will fit the word length of
the computer being used. For
example, in an eight-bit com-
puter the complement of 5
would be 11111010 binary,
or 372 octal.
So, whenever you find the
complement of any number,
always be sure to keep in
mind the word length of your
computer, and modify the
answer to fit your word
length. In the case of hobby
computers, this problem
usually arises on either the
Heath H8 computer or any
8008 system, which use octal
with an eight-bit word length.
Since the leftmost octal digit
of any octal number on these
computers only stands for
two binary digits, the largest
it can be is octal 3 (or binary
11). Hence, any complement
that starts with a digit greater
than 3 is wrong. The usual
trick is to subtract a 4 from
the leftmost digit.
Suppose you want the
one's complement of 005. If
you follow the rule for con-
verting, you get
777
-005
772
Since the leftmost digit is
greater than 3, there is an
extra bit. Remove it by sub-
tracting 4 from it, so the
actual complement is 372.
Converting a hex number
to one's complement. The
rule is the same as for octal
numbers, except that we
write a 15 above each digit
and convert hex digits to and
from decimal.
The one's complement of
hex 68 is hex 97.
15
-6
15
-8
The one's complement of hex
9E is hex 61; we have to
convert E to 14:
15
-9
15
- E (14)
1
The one's complement of hex
61 is hex 9E; this time we
Binary Number
Binary Complement
Octal Complement
101
010
2
0101
1010
12
00101
11010
32
000101
111010
72
Fig. 1.
have to convert 14 to a hex
E:
Write the binary number.
Now find the rightmost 1 and
put a vertical line just to the
left of it. Invert all bits to the
left of this line. Leave the bits
to the right of the line un-
changed.
Convert the binary num-
ber 10110 thus:
1
0 1
1 0
15
15
0
1 0
1 0
-_6
-_±
invert
leave
9
14(E)
alone
The same warnings about
extra ones in the complement
apply here as when using
octal numbers; but we don't
usually have to worry about
it because in most computer
systems the number of bits
matches the hex digits ex-
actly. For example, the two
hex digits used in eight-bit
computers like the 8080 or
6800 match the word length
exactly.
Converting numbers in
one's complement to two's
complement. As mentioned
before, most systems use
two's rather than one's
complements. It's easy to
convert from one's to two's
complement: add 1. If the
one's complement of some
number is 110, the two's
complement is 111; if it's 61 ,
the two's complement is 62;
if it's 9B, the two's comple-
ment is 9C — adding 1 to B
(which is 1 1 ) makes it C (1 2).
Be careful how you add 1
— it has to be done right. For
example, if the one's comple-
ment is a binary 101, adding
1 does not give you 1 02
because a 2 is not allowed in
binary! 101 plus 1 is 110
(refer to the table).
Although this is irrelevant
anyway since there are other
ways of converting, it is of
some interest since many
microprocessors convert to
the two's complement by
first finding the one's comple-
ment and then adding a 1.
For instance, the Intersil
6100 has a CIA (complement
and increment accumulator)
instruction. (Increment
means to add one.)
Converting a binary num-
ber to its two's complement.
The two's complement of the
eight-bit number 00000101 is
11111011 :
0 0 0 0 0 1 0
111110 1
invert
Converting an octal num-
ber to its two's complement.
Write the octal number and
see whether it has any zeros
at its right end (ignore zeros
in the middle or at the left).
If so, put a zero above each
zero at the right. For in-
stance, if you wanted to con-
vert the octal number
02307500, you would now
have
0 0
02307500
Continue from the right and
put an 8 above the next digit
and a 7 above each of the
others. Finally, subtract each
digit from the one above it
(see Example 2). The two's
complement in this case is
75470300.
Just one warning: Every-
thing we said about extra
ones in the one's complement
conversion applies here, too.
For instance, in an eight-bit
computer the complement of
005 would be 373, not 773.
If you find this method
too hard to remember, you
can always convert your octal
number to binary, find the
two's complement of that,
and then convert that back to
octal.
Converting a hex number
to its two's complement.
Look at the hex number to
see whether it has any zeros
on the right end (ignore zeros
in the middle or at the left).
If it does, put a zero above
51
15
15
16
0
- C (12)
-_0
-B (11)
-9_
3
15(F)
5
0
Example 3.
each of these rightmost zeros.
To convert COBO, you would
write:
0
C 0 B 0
Continue from the right and
write the number 16 above
the rightmost nonzero digit
of the hex number; write 15
above each of the other
digits. Finally, subtract each
of the hex digits from the
number above it, converting
from letters to numbers — or
back if needed. COBO con-
verts to 3F50 (Example 3).
As another example, the
two's complement of hex 05
is FB:
15 16
-_0 -_5
15 (F) 11 (B)
By the way, the two's
complement of a two's
complement is the original
number; the two's comple-
ment of FB is 05:
15 16
-F (15) -B (11)
0 5
Converting Decimal to BCD
Many computers allow
calculations to be done in
binary coded decimal (BCD)
rather than only in binary.
(BCD is a combination of
binary and decimal.) Con-
verting decimal to BCD is
performed in the same way as
converting hex to binary:
Replace each decimal digit by
its four-bit binary equivalent
from the table. To convert
decimal 93, replace 9 by
1001 and 3 by 0011 to get
10010011 .
Notice that this result is
different from the 01011101
you would get if you con-
verted 93 to binary. In con-
verting to binary, you convert
an entire decimal number at
once; in converting to BCD,
you convert only one digit at
a time.
Watch out for one big area
of confusion. If you convert
decimal 93 to BCD you get
10010011, which looks like
binary. Consequently, you
might be tempted to convert
this "binary" number to hex,
by following the standard
procedure, to get 93.
This might fool you into
thinking that hex 93 is the
same as decimal 93, which is
not so. The "hex" 93 is not a
true hexadecimal number; it
is only a form of shorthand
that allows you to express the
bit pattern 10010011 in a
simpler form. If you were
employing an assembler that
used hex, you might use what
looks like hex 93 when you
really meant BCD 10010011.
BCD to Decimal
This conversion is the
same as that for binary to
hex: Arrange the bits in
groups of four starting from
the right, and convert each
group into hex using the
table. For instance, BCD
10001001 is grouped into
1000 and 1001, which gives
the decimal 89.
In BCD to decimal, you
should never get the digits A
through F. If you do, then
the BCD number was wrong.
For instance, to convert
00111100, you would get
two groups 0011 and 1100.
The 001 1 converts into a 3,
but 1100 converts to C,
which is not allowed in
decimal. Hence, 00111100
was not a valid BCD number.
So - What's All This Used
For?
If all your programming is
in BASIC, you will probably
never need to know any of
this hex magic. But if you do
any machine- or assembly-
language programming, it will
help a lot.
For example, suppose you
want to set up a counter at
-50 (decimal) and want to
convert this to hex. First find
+50 in hex: 50 divided by 16
is 3, with a remainder of 2; 3
divided by 16 is 0, with a
remainder of 3. So, a decimal
+50 is hex 32. Now change
this to -50 by finding the
two's complement:
15 16
-_3 -_2
12(C) 14(E)
-50 is CE in hex.
Or suppose you want to
subtract 2 from some hex
number. If your computer
does not have a subtract in-
struction, you can do the
same thing by adding a -2. In
hex, 2 is 02, and the -2 is
found as the two's comple-
ment:
15 16
-_0 -_2
15(F) 14(E)
You should add hex FE.
Once you figure it out,
hex magic can be fun. ■
16K
S-100
STATIC RAM
*295
• FULLY STATIC (Not Pseudo Static)
• PHANTOM DISABLE (A16)
• MWRT OR PWR
• FULLY BUFFERED
• 128-2102L1PC S
• 400NS GUARANTEED (250NS TYP.)
• ASSEMBLED/BURNT-IN
• 90-DAY GUARANTEE
Also:
8K STATIC RAM *155
2102L1PC *1.40
Calif. Residents add 6%. • Master Charge & Visa welcome
P.0. BOX 17296
IRVINE, CA. 92713
G15
(714) 751-7341
RO-CHE Systems
MULTI-CASSETTE
CONTROLLER
• Read and write records from and to up to 4
cassette recorders with one Tarbell Cassette
Interface.
• Included software handles Assembly Lan-
guage and BASIC.
• File Maintenance System and Text Editor
available.
Write for brochure:
RO-CHE Systems R16
7101 Mammoth Avenue
Van Nuys, California 91405
Canadian
8K MEMORY KITS
• Low Power, 500NS, S 1 0O BUS
• On-Board Regulation, No Duty
• Prime Quality, First Run 21 LO 2 ICs
• WAMECO PC Board, SST Included
• Full Documentation, Tl Sockets
• Solder Mask
Price . . . $219.95 (Canadian)
IchxIvisaI
CHARGEX & VISA
HONORED
Mail orders to: /W TnC)j nc
ORTHON COMPUTERS
12411 Stony Plain Rd
Edmonton, Alberta Canada T5N3N3
08
52
By subscribing to COMPUTER DIGEST, you
have the facts you need about computers and
optical scanning.
With this information, you can act immediate-
ly on new business opportunities. Subscribe and
each month you get:
• A 12-page newsletter digesting the im-
portant news in the computer and optical scan-
ning field. There is coverage of new equipment,
new optical scanning and computer forms, who
is expanding, who is getting promoted— a wide
variety of facts, all written in concise style, so
you can read the reports in minutes.
• Over 50 news reports in each issue cov-
ering new equipment, new uses of computers
and scanning equipment, new products and
software.
the
• Not only can you find new markets, but
reports will help you solve customer
problems, because you will be more knowledge-
able about computers and optical scanning.
• It’s a timely summary of what is being
written about computers and optical scanning in
the business press. You save reading time, plus
you can be sure of getting hundreds of new
items of interest each year.
To subscribe, just complete the form below
and mail it back. The newsletter is fully guaran-
teed. Read 2 issues, see if you can use the infor-
mation. If you find you don’t need COMPUTER
DIGEST, just write “cancel” on the invoice we
send you, and owe nothing. If you want to con-
tinue to read COMPUTER DIGEST, simply pay
the $40 invoice we send. We have found this is
the best way to introduce businessmen to this
newsletter. This way, we take all the risk, so you
have nothing to lose by trying this publication.
CLIP AND MAIL
Send to:
COMPUTER DIGEST
North American Building, 401 N. Broad St., Philadelphia, PA 19108
Yes
send me COMPUTER DIGEST, also incorporating OPTICAL
y SCANNING NEWS. I understand I can look at 2 issues, and if not
satisfied, I will owe nothing. I can write “cancel” on the bill, and
owe nothing. If I like the newsletter, I will pay the $40 bill I receive.
□ Bill company for $40
□ Bill me for $40
□ Payment enclosed
Name-
- Title.
Company -
Address —
City/State/Zip
% | Computer Digest is published by North American Publishing Company, leaders in editorial excellence.
KB/5/78
53
Bus Traffic Control
George Young
Sierra High School
Tollhouse CA 93667
I n the last session, we covered
the majority of the TTL
counters and some of the regis-
ter chips. We performed many
experiments with these chips,
thus building your background
skills in reading circuit dia-
grams and, I hope, building up
your confidence as well. As you
can see, the sessions are be-
ginning to get a bit rougher.
Hang in there; we will make it
yet.
In this session, we will take
up decoders, decoding, three-
state devices, and how traffic is
controlled on the microproces-
sor data bus.
Introduction
Most of our modern micro-
processor chips have 16 ad-
dress lines providing the capa-
bility of selecting 65,535 dis-
crete memory locations. These
separate memory locations are
referred to as the address
space of the microprocessor.
Fig. 1 shows the microproces-
sor and 16 address lines. These
are labeled A 0 through A 15 . 1 K of
RAM requires ten address lines
from the microprocessor to
select the 1024 separate mem-
ory cells in each RAM chip; so
we have drawn the 10 address
lines A 0 through A 9 running
from the microprocessor to the
IK RAM block.
We are going to draw the IK
RAM block in an unusual
fashion. There are actually
eight separate RAM chips in
the RAM block, and we have
drawn them stacked up in order
to conserve space. We did not
draw eight rectangles in the
stack, but the concept of more
than one chip is readily con-
veyed by this diagram.
As shown in Fig. 1, our IK of
RAM will not function; two
things are wrong. First, the ten
address lines will not drive the
address inputs of the IK RAM
block. Microprocessor output
pins are capable of driving one
TTL load. We are asking each
address line to drive eight in-
puts to the RAM block. There-
fore, we must provide buffering
on each of the address lines out
of the microprocessor. A buffer
is a circuit placed between two
circuits to provide isolation. We
need a buffer on each address
line not for isolation, but to in-
crease the drive capability.
The second reason Fig. 1
won’t work is that the chip en-
able (CE) pins on the RAM
chips are floating. The CE and
the small circle on the symbol
both indicate that we need an
active low enable here to make
the RAM function.
In Fig. 2a, we have added
noninverting buffers to each
address lineto provide the drive
capability required. Our first
idea is to use the A 10 address
line forthe"CE input forthe first
IK RAM block. After all, this
line will be low for the first 1 K of
memory space; and when this
line goes high, the first IK RAM
block will be de-selected.
We are also introducing
another concept in Fig. 2a. The
ten address lines, A 0 through
A 9 , are shown entering a rec-
tangle. Feeding from the rec-
tangle is a widened arrow that
goes to each of the RAM
blocks. Data lines and address
lines are often drawn in this
fashion. The broadened line in-
dicates that more than one line
is included in the wide line. This
saves drawing the individual
lines involved and takes less
space in the diagram. As long
as the idea is understood by
everyone, there is no problem,
and the diagram is clearer and
actually more easily under-
stood.
Furthermore, in Fig. 2a we
have added a second IK RAM
block. Our first thought on
Fig. 1. Addressing the IK RAM block.
54
MICROPROCESSOR
MICROPROCESSOR
7 i iiii tiii
Y
7 -
AI5 AI4 AI3 AI2 All AIO A9 A8 A7 A6 AS A4 AS A2 Al AO
32K 8K 2048
<$-
CE-2
1 &—
^ CE-3
CE-4
#-
LED ON ■ CHIP ENA8LE
Fig. 3. Experimental setup for decode testing.
ADDRESSES
AI5
AI4
Al 3
Al 2
All
AIO
4- 1023
0
0
0
0
0
0
1ST
IK
1024-2047
0
0
0
0
0
1
2ND
IK
2048-3071
0
0
0
0
1
0
3RD
IK
3072-4095
0
0
0
0
1
1
4TH
IK
4096-5119
0
0
0
1
0
0
5TH
IK
5120-6143
0
0
0
1
0
1
6TH
IK
(b)
Fig. 2. Adding the second IK RAM block.
handling the second group of
UE pins on this block is to add
an inverter between the'CE on
the first RAM block and the~CE
on the second RAM block. This
will work if we only have 2K of
memory in our system. If we
have more RAM or ROM, then
an examination of the truth
table in Fig. 2b will help us find
out why this simple method of
enabling the 2K will not work.
The truth table shows that
the A 10 line does indeed start
out low for the first IK of
memory space and then is high
for the second IK. But lines 3
and 5 of the table also show the
A 10 line low. Therefore, the first
IK RAM block will be selected
every time the A 10 line goes low.
In other words, the single in-
verter decoder will not do for
memory sizes above 2K.
Fig. 3 shows the experimental
setup for the design console
breadboard and the address
lines from the microprocessor.
Since we don’t have a micropro-
cessor (yet), we’ll use this cir-
cuit to show how the lines are
related; the actual test circuit is
shown in Fig. 4a. The chip en-
able LEDs have been arranged
in the circuit to turn on the LED
when the CE line goes low.
In Fig. 4 we are attempting to
place an equivalent circuit on
the console breadboard that
will represent what happens
with the address lines and the
decoding process. Fig. 4a
shows the equivalent bread-
board circuit for Fig. 2. Note
that we are not considering the
A 0 through A 9 address lines in
the decoding process. These
lines are used by each IK block
of memory throughout the ad-
dress space and are not used in
the decoding process for each
IK block.
In order to have a IK RAM
block selected only once in the
memory space, we must use
some form of decoding. We can
use gates and inverters and de-
code each IK block in this
fashion. Fig. 4b shows this kind
of decoder. You can set this cir-
cuit up on the console bread-
board and use it to decode the
four CE lines; but there is an
easier way— use a decoder
chip. This makes a rather long
introduction, but I think that we
have the problem fairly well
delimited.
Experiment #51
The 7442 Decoder
Problem: How can the ad-
dress lines of the microproces-
sor decode the memory chips?
Solution: We will investigate
this on the console breadboard.
♦5V
The experiment uses the 7442
decoder, but the 7441 , the 7445,
the 74145, the 8250 or the 8251
may also be used for this ex-
periment.
Procedure: Refer to Fig. 5.
Fig. 5a shows the 7442 pin-
outs; Fig. 5b shows the 7442
truth table. Notice the row of
zeros (lows) traveling diagonal-
ly across the truth table . . . this
is exactly what we need for chip
enable pins. Put the 7442 on the
console breadboard (don’t for-
get power and ground). Use
four jumper wires to represent
the A 10 through A 13 address
lines. Start with all four inputs
to the 7442 grounded. The LED
marked CE-1 should be on.
Theory: The 7442 is a one-of-
ten (usually written 1:10) de-
coder. It has four input lines
marked A, B, C and D on our dia-
gram. The truth table of Fig. 5b
Fig. 4. Delimiting the address decoding problem.
55
shows that with all inputs low,
the 0 output line (pin 1) will be
low. This should turn on CE-1.
This line would, therefore, go to
the first IK RAM block "CE pin,
and would select that RAM
block. (Fig. 5c illustrates
decoding 4K of RAM.)
Now take the A 10 jumper wire
high. This should turn on CE-2
and turn off CE-1. This line
(from pin 2 on the 7442) would
go to the second IK RAM block
and select this RAM block
while, at the same time, the
first IK RAM block is de-
selected.
If you now encode a binary 2
by taking the A„ line high and
the A 10 line low, pin 3 on the
7442 should go low, turning on
CE-3 and turning off CE-2. This
line from pin 3 on the 7442
would go to the third IK RAM
block and select it while blocks
1 and 2 are de-selected.
Finally, if you encode a
binary 3 with both the A 10 and
A„ lines high, CE-4 will illumi-
nate and CE-3 will turn off. Pin 4
of the 7442 would go to the
fourth IK RAM block selecting
it while the highs on pins 1, 2
and 3 will de-select the first
three RAM blocks. Thus, we
have a decoder for 4K of mem-
ory chips.
But wait, we did not use all
the outputs of the 7442. What
about the rest of the output
pins?
The 7442 may be operated as
a 1:4 decoder, 1:8 decoder or
1:10 decoder. To use only the
first eight outputs of the 7442,
we do not use the D input to the
7442; we leave it grounded. We
can then operate the 7442 as a
1:8 decoder and use the eight
output pins to decode 8K of
RAM. To operate the 7442 as a
1 :4 decoder as we just did in the
experiment, leave the C and D
inputs grounded and operate
the 7442 as a 1:4 decoder to
decode 4K of address space.
We may use all ten out pins of
the 7442 and decode 10K of ad-
dress space with the 7442.
Fig. 6 gives the pin-outs for
several more decoder chips.
Experiment #52
The 74154 Decoder Chip
Problem: To decode more
than 10K of address space.
Solution: Use a decoder that
has more output pins.
Procedure: Refer to Fig. 6e,
where the 74154 1:16 decoder is
set up in a test circuit. This
24-pin chip was designed for
address decoding in com-
puters. It has two enable pins,
18 and 19. Use two jumper wires
on these pins to represent the
A 14 and A 15 address lines. Any
binary counter may be used to
simulate the A 10 through A 13 ad-
dress lines. Set up the circuit
with the 74161 counter chip.
Sixteen LEDs are shown moni-
toring the 74154 output lines.
If you do not have 16 LEDs,
then use as many as you can
for the test circuit. Remember
that the console logic probe
may be used for one LED and
that you have eight LEDs in the
console 7-segment readout. If
you have the FND 70 readout,
then it will be necessary to
drive the segments of the FND
70 through inverter sections
since FND 70 requires an active
high to turn on each segment.
The 74154 will decode 16K of
2
VCC
J6
25
3
O A
14
*
5
1 „
L
2 C
3 D
_I3
12
6
«— c
4 9
3 8
7
8
1 — c
GND
6 7
10
9
(o) 7442 PIN-OUTS
address space.
Experiment #53
The Traffic Cops
Problem: What is all this stuff
hung on the data bus lines?
Solution: Let’s take a look.
Procedure: Fig. 7 shows the
microprocessor chip and its
eight data lines. It also shows
arrows signifying data travel-
ing both directions on these
data lines. During a read cycle,
the data is traveling from mem-
ory (or input/output devices) in-
to the microprocessor. During a
write cycle, data travels from
the microprocessor out to ex-
ternal devices. Fig. 7b shows a
single data line (D 0 ) and a pair
of open collector NAND gates
acting as traffic cops on the
data line.
Theory: Assume that the
microprocessor is in a memory
read cycle. This means that the
R/W is high. The high on pin 2
of the 7403 will enable this
gate, which means the data to
be read into the processor will
be enabled. This high is also in-
verted to a low by the inverter
INPUTS
OUTPUTS
0
c
B
A
0
1
2
l
4
5
6
7
8
9
0
0
0
0
0
1
1
1
1
1
1
1
1
1
0
0
0
1
1
0
1
1
1
1
1
1
1
1
0
0
1
0
1
1
0
1
1
1
1
1
1
1
0
0
1
1
1
1
1
0
1
1
1
1
1
1
0
1
0
0
1
1
1
1
0
1
1
1
1
1
0
1
0
1
1
1
1
1
1
0
1
1
1
1
0
1
1
0
1
1
1
1
1
1
O
1
1
1
0
1
1
1
1
1
1
1
1
1
1
O
1
1
1
0
0
0
1
1
1
1
1
1
1
1
O
1
1
0
0
1
1
1
1
1
1
1
1
1
1
O
1
0
1
0
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
0
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
(b) 7442 TRUTH TABLE
♦5V
Fig. 5. The 7442 decoder.
section, and the low on pin 5
will disable this gate (taking it
out of the circuit for the time
being).
Next, the microprocessor is
assumed to go into a memory
write cycle. The R/W goes low,
and the low on pin 2 of the 7403
now disables this gate and pin
3 floats on the end of the 2.2k
pull-up resistor. The low on the
R/W line is inverted by the in-
verter section, and the result-
ing high output applied to pin 5
will enable this gate. The data
to be written into memory (from
the processor) will now be en-
abled onto the data bus. This
circuit illustrates how the two-
way traffic on the data bus is
controlled by the “traffic cops”
in the circuit. The R/W line and
the inverter control the two
gates and the direction of the
traffic flow.
Fig. 7b is fine for an introduc-
tion and example of controlling
data on a bus going to and from
the processor. However, it isn’t
practical from a design stand-
point (for several reasons).
First, the dual-gate configura-
tion would have to be repeated
for each data line. This means
the R/W output from the micro-
processor would be driving
eight gates. You’ll recall from
an earlier discussion that all of
the microprocessor outputs are
capable of driving only one TTL
gate each. Fig. 7c illustrates a
solution to the problem— the
addition of an inverter, and a lit-
tle reconfiguring. Now the R/W
signal is going into the 7404
(pin 1), which is driving the
eight write gates (only one of
which is shown).
The second, and most impor-
tant, reason why this circuit is
totally unacceptable lies in the
use of the 7403 gates for inter-
facing with the bus. The whole
idea behind a bus system is
that several devices can be
plugged into the bus (i.e., other
gates will be tied to the bus fur-
ther down the line). These addi-
tional gates have a “loading ef-
fect” on the bus. Without my
going into a detailed technical
explanation, it will suffice to
say that such systems con-
sume a lot of power and are
noisy (i.e., have glitches and
56
spikes that can be interpreted
as logic ones or zeros). The
answer to the problem is to use
Tri-state gates for interfacing
to a bus.
Tri-state gates, such as the
8T97 shown in Fig. 8a, are
either enabled or disabled.
When they are enabled by a low
on the Disable pins (1 and 15),
the outputs will be determined
by the logic levels (HI or LO) at
the input pins. In other words,
the gates are working just like
any other gates. When they are
disabled (by a high on the Dis-
able line) the gates are effec-
tively disconnected from the
bus. The outputs are said to
have gone into a high-imped-
ance or open condition and do
not present any loading to the
bus (i.e., they are discon-
nected). Fig. 8b is a truth table
for the operation of the 8T97
and Fig. 8c illustrates a typical
bus interface configuration.
In summary, there are three
advantages to using Tri-state
gates when you are interfacing
to a microcomputer bus (one of
which I haven’t mentioned be-
fore). First, lower power con-
sumption; second, less loading
on the bus (thereby maintaining
waveform integrity); and finally,
higher speed (faster switching
from a high to low or vice versa).
Note that the 8T97 is a nonin-
verting buffer and has four sec-
tions controlled by one line and
two sections controlled by a
second line. The two sections
may be operated independently
of each other. The DM 8097 and
the 74367 are also the same
type of chip. The 8T97 is more
CONSOLE
CLOCK
IN
Fig. 6. The 74154 decoder.
Fig. 7. Traffic control on the data bus.
57
expensive than the others, but
my own experience with these
chips indicates that the 8T97
has more drive capabilities and
proves superior in operation in
the circuit . . . justifying its
greater cost.
Other chips are becoming
available for this buffering job
on the data and address buses;
I think that soon we may see a
new family of microprocessors
with the buffers, as well as
RAM and ROM, built into the
chip. In fact, Intel has a new
microprocessor chip, with
many of these capabilities built
in, which will be second-
sourced by Signetics. This
points the way that things are
heading in the subsequent gen-
eration of microprocessor
chips.
Preview
We have looked at the micro-
processor address bus, how de-
coding of the address space
may be accomplished and how
traffic is controlled on the data
bus.
DISABLE
DIS2 DIS4
INPUT
OUTPUT
0 0
0
0
0 0
1
I
0 1
X
HI Z
1 0
X
HI Z
1 1
X
HI Z
(b) TRUTH TABLE
(a) 8T97 PIN-OUTS
DM8097
74367
Next time we will turn our at-
tention to the memory chips,
both ROM and RAM. Using the
7489 (8225), we will set up 64
bytes of memory on the con-
sole breadboard, and also burn
a 7488 (8223) PROM on the con-
sole. Sierra Electronics, Box 11,
Auberry CA 93602, will furnish a
package for us of two 8225s
and two 8223s for $4 postpaid
in the U.S. and Canada. Califor-
nia residents, add 6 percent
sales tax.B
Fig. 8. 8T97 data bus buffering.
We speak your
language
And we’re giving you what you want.
• a comprehensive product line. Hardware, assembled or
kits, and software from major manufacturers. Plus books and
current literature. Financing available.
• a trained, enthusiastic staff. We’ll help you choose or
design the system that’s right for you . No high pressu re here.
• service when you need it. We won’t sell you something
we can’t keep running.,^
• a brand new facility
in Pennsylvania.
We’ll be able to
serve our South Jersey-
Pennsylvania customers
more efficiently now.
• a bigger, better New
Jersey store. We’ve
enlarged our showroom
in Iselin. Now there are
more displays you can
try out. There’s more
room to stock the
products you need.
The Microcomputer People . ®
Computer Mart of New Jersey
Computer Mart of Pennsylvania
C30
New Jersey Store
501 Route 27
Iselin, NJ 08830
201-283-0600
Tue.-Sat. 10:00-6:00
Tue. & Thur. til 9:00
Pennsylvania Store
550 DeKalb Pike
King of Prussia, PA 19406
215-265-2580
Tue. -Thur. 11:00-9:00
Fri. & Sat. 70:00-6:00
(our only locations)
INNOVEX FLOPPY DISK DRIVE
NEW-FULL SIZE $495 or LESS*
From The Same People Who Brought You The First Diskette and Drive
TO DESIGN YOUR OWN CUSTOM SUB-SYSTEM FOR LESS:
( 1 ) BUY THE INNOVEX MODEL 41 0 OR 420 FROM THE MANUFACTURER
AT THE OEM VOLUME PRICE OF $495.
8" Full Size, IBM Compatible,Hard
or Soft Sector Single or Double Density
(2) BUY THE FLOPPY CONTROLLER DESIGNED FOR YOUR SYSTEM
(3) THEN FOR A LIMITED TIME GET A REBATE FROM INNOTRONICS
HAVING BOUGHT ONE OF THESE POPULAR INTERFACES SPECIFI-
CALLY DESIGNED TO WORK WITH THIS INNOVEX DRIVE:
Tarbell $20 Peripheral Vision $10
SCI $20 Digital Group $10
WE ARE PROUD TO OFFER THE HIGHEST QUALITY DISKETTE DRIVE
FOR THE FAIREST PRICE POSSIBLE IN THE INDUSTRY
ACT NOW AND SAVE!!!!
Send Check Or Money Order for $495
Less Rebate If Applicable
(proof of purchase required)
INNOTRONICS CORPORATION/DEPT A
BROOKS ROAD
LINCOLN, MASSACHUSETTS 01773
Tel. (617) 259-0600
* Offer Ends 7-1-78
58
The many faces of MERLIN
ftinlTera HERLIM
Dan*# (ICiH bU
in tha fiix*4,|
188 V > graphic nod*.
-V
mi
sir \y
k
^ Jfc hjl.
•
Dense Mode: 160H * 100V
Running Man Patterns
Monitor Debug Usage
Super Dense: 320H * 200V
3-D Boxes
Super Dense: 320H * 200V
Line Drawing
Super Dense: 320H * 200V
Equation Plotting
3882
REK/ MAIM 1
LOOP
3884
CALL S3 -REM/ EDI8 (FLIP SCREEN)
3121
1-1
3839
P8»i3DB9
3835
COSUB 2328
3848
G0SUB 4828
> REN/
CALC PATTERN
3838
COSUB 2228
* REN/
0RAH PATTERN
3848
P8»P8*Ii98
3865
COSUB 2528
3878
COSUB 4828
• REH/
CALC NEXT PATTERN
3888
C0SU8 2928
■REN/
DELAY
3898
P8»P8~IX98
3188
COSUB 2326
> REH'
CLEAR
3118
3128
P8*P8-I»98
3138
GOTO 3658
4818
REH/ C ALUL ATE PATTERN SOB
4828
H*28+RND<Z)X88
4838
L*28+RND< 2)488
4848
D*18*RND< Z >X58
BASIC Program Listing
Output Shown Below
Super Dense: 320H * 200V
Perspective Drawing
MERLIN (and your S-100 Computer...)
... the graphics development package you’ve been waiting for.
Do these photos suggest an application in your field? Whether you’re into architecture, astro-
logy, music or whatever, you can apply graphics to make your presentations more effective
and your work more efficient.
Real Time Plotting
Heart Rate
Navigation
Spectrum Analysis
Complex Equation Plotting
Stock Market Trends
Teaching: Mathematics
Circuit Responses
Fine Line Drawings
Architecture
3-D Projections
Circuit Layouts
Pattern Movement
Animation
Games
MERLIN is also a multipurpose system monitor board with
ROM monitor and editor software, parallel keyboard port, and
audio cassetts storage, besides being your text and graphics
video output device.
Write or call today for your free copy of our new catalog, an
index to available application notes, and a list of MiniTerm
Dealers. ^
MERLIN is supported by expert technical assistance, a con-
stantly expanding series of application notes, and a newly
formed users group.
Assembled and tested MERLINs start at $349, $269 for kits.
Super Dense add-on, firmware, and cassette interface are extra.
MiniTerm Associates, inc.
Dundee Park, Andover, MA 01810 (617) 470-0525
M40
Expand Your KIM
Part 5: A/D interfacing (for joysticks!)
i :
ii
liiS
2 : : : :
Photo 1. Four channels of A/D, two channels of D/A and an input
port for sense switches.
mat
nm
|&||
w 1
1 < ‘
* H
John Blankenship
data mart, inc.
3001 No. Fulton Dr. N.E.
Atlanta GA 30305
N o matter what kind of com-
puter you have, this article
can help you add four channels
of analog input for a fraction of
the cost of other methods I’ve
seen. If you’ve been building
the KIM System, this analog
board will complete the project.
I designed the KIM System
with several requirements in
mind for the analog ports: I re-
quired four channels (so that
two joysticks could be inter-
CLOCK JUUL
8-BIT
COUNTER
END OF
COUNT
y\y\yv
■>
4 WORD
8- BIT/ WORD
MEMORY
0
a
SYSTEM l
PROCESSOR I
MEMORY WRITE
PULSE
DIGITAL -
TO-ANALOG
CONVERTER
TWO-BIT
COUNTER
ii
r~
2 IN 4 OUT
DECODER
CONTROL
DATA
EDGE
DETECTOR
4 TO I
OATA
MULTIPLEXER
Photo 2. Circuits are wire-wrapped on a 44-pin board.
Fig. 1. Block diagram of the A/D converter.
60
faced), with each sampled
often enough to provide rea-
sonable accuracy for use as
a video game input device. To
make the use of these ports
easy, I wanted each to be read
as a normal memory. Finally,
each of the A/D (analog-to-digi-
tal) channels had to be easily
switchable to other devices be-
sides the joysticks.
Besides the A/D ports, I also
wanted at least two D/A (digital-
to-analog) ports to experiment
with music, speech synthesis,
motor control, etc. I also
wanted a port for sense switch-
es to give me a full complement
of methods for interfacing with
my machine. I combined all
these circuits on one board and
labeled it External Interfacing
in my previous articles.
Photos 1 and 2 show the
board itself. Although I was
able to cram the circuit onto a
4V2-inch-square board, I would
recommend epoxying a vector
board on the top to give more
room for the components.
Fig. 1 shows the basic block
diagram for the A/D circuits.
The four-word memory is one of
the major secrets of making
this circuitry both inexpensive
and easy to use. This memory is
made up of two 74LS170 chips
composed of four 4-bit words
each. I chose these chips be-
cause they have separate read
and write controls, thus en-
abling read and write opera-
tions to occur simultaneously.
The A/D circuitry will update
each of these memory loca-
tions with a number that is pro-
portional to the analog input.
The output of the memory chips
is connected to the data bus so
that they appear as standard
memory to the processor.
The eight-bit counter contin-
ually generates sequential
numbers from 0 to 255. A D/A
converter converts these num-
bers to an analog voltage
which, for all practical pur-
poses, is an increasing ramp.
This ramp is fed to four com-
parator circuits that compare
the ramp voltage to the analog
inputs.
The comparators output a
level 1 when the ramp voltage
equals the analog input. Since
the ramp voltage also equals
the number in the eight-bit
counter, it is implied that the in-
stant a comparator fires, the
eight-bit counter contains the
digital equivalent of the analog
voltage being applied to that
comparator.
The remainder of the circuit
has one major function ... it
must decide which comparator
fired, and form an address for
the four-word memory so the
eight-bit counter data can be
gated into the appropriate lo-
cation.
I chose to control the write
address with a two-bit counter.
Since this counter increments
every time the eight-bit counter
completes a full cycle, the ad-
dresses 0, 1, 2 and 3 are being
applied sequentially to the
write address, and each is held
there for the full cycle of the
eight-bit counter.
Additionally, this two-bit
counter is decoded and used to
enable only one of the four
comparators (the one corre-
sponding to the write address)
at a time. The level change in-
dication from the multiplexer is
converted to a narrow pulse
and used to activate the write
line on the memory chips.
As explained above, the four
memory locations are continu-
ally, and automatically, re-
freshed with the digital equiva-
lent of four analog inputs. The
processor needs only to read
these locations for the latest
updates.
Fig. 2 shows the actual sche-
matic of the A/D circuit. The
7493 simply reduces the fre-
quency to a trackable rate. The
1408L8, D/A converter, outputs
a current ramp that is convert-
ed to a voltage ramp by the 741
♦5
Fig. 2. Schematic of A/D converters.
61
( 9 ) ( 13 )
RAM R/W I/O ENABLE
Fig. 3. Schematic of D/A converters and input port for sense switches.
op amp. The 7400 labeled B
acts as a one-shot to perform
as the edge detector.
Half of the 7420 is used to de-
code the address bus for pro-
cessor reads. Address decod-
ing will be discussed in more
detail later in this article.
Since the 74LS170s are open
collector, rather than Tri-state
outputs, pull-up resistors are
required for interfacing with the
bus. The DIP switch discon-
nects the joystick inputs. Once
they’re disconnected, you can
input other signals to the con-
verter by way of the backplane
jacks (see my earlier articles).
The other two functions, D/A
and sense switches, are de-
tailed in Fig. 3. Since I felt that
Port Function
Dazzler Mode control
Dazzler ON/OFF, Address
Right vertical joystick
Right horizontal joystick
Left vertical joystick
Left horizontal joystick
Sense switches
D/A port A
D/A port B
the accuracy of the D/A conver-
sion was not critical, I chose
not to use the Motorola D/A
converter chip used in the A/D
circuit. If I had used the Moto-
rola chip, I would have had to
use two eight-bit registers to
hold the data, the two D/A chips
themselves and a current-to-
voltage converter.
I chose to use MOS registers
for my output ports. Since MOS
gates output exactly Vcc and
zero volts for their correspond-
ing high and low levels, I used
them to drive a resistive ladder
directly. Additionally, since
MOS chips represent a very
small load, they can be hung on
the bus without buffering.
(Note: MOS chips do represent
Page
Loc
80
OF
80
0E
80
10
80
11
80
12
80
13
80
80
80
20
80
40
a relatively large capacitive
load, and hanging them directly
on the bus is not good practice
in expandable systems. In this
case, however, I knew exactly
what loads I would be dealing
with and was able to determine
that enough drive capability
was present.)
The 741s in Fig. 3 are used as
A/D D A
A/DC B
A/D B C
A/D A D
D/A B E
D/A A F
-12 H
JS REF. Volt. J
02 K
Ground L
SS 0 M
SS 1 N
SS 2 P
SS 3 R
SS 4 S
SS 5 T
SS 6 U
SS 7 V
JS LH W
JS LV X
JS RH Y
JS RV Z
unit gain amplifiers for buffer-
ing purposes. The 7410 is used
to decode out the address lines
to determine which port is
being used. The sense switch-
es are connected to the inputs
of Tri-state buffers. The out-
puts of these buffers gate the
switch data onto the bus when
enabled.
1
ABO
2
AB1
3
AB2
4
AB3
5
AB4
6
AB5
7
AB6
8
AB7
9
RAM R/W
10
+ 12
11
12
+ 5
13
I/O ENABLE
14
W/R
15
DB7
16
DB6
17
DB5
18
DB4
19
DB3
20
DB2
21
DB1
22
DB0
Fig. 4. Summary of special addresses used by the KIM-1 System. Fig. 5. Pin-out designations for the external interface board.
62
In order to better understand
the I/O functions, you might re-
read my article (“Expand Your
KIM!” Part 3, Kilobaud, Feb-
ruary 1978, p. 68) in which I ex-
plain how I decoded part of the
address lines to indicate an I/O
operation, rather than a mem-
ory transfer.
All my I/O ports (including
the four-word memory used for
A/D) are partially enabled by
this I/O enable. Since I know
how many total ports I de-
signed for, I only partially
decoded the low-order address
lines. This drastically limited
the number of ports available
on the KIM System, but the
ease of implementation, as well
as the reduction in cost, made
it well worthwhile.
Fig. 4 summarizes the I/O ad-
dresses used uniquely by my
system. If you convert these
hex addresses to binary, you
can see how the appropriate
address lines are used to en-
able each port decoder.
There are only two major dif-
ferences between input and
output decoding. The first is
that the R/W or the R/W line is
used to indicatethedirectionof
the transfer. Second, the write
pulse for an output port must
be coincident with the trailing
edge of the 02 clock. Again, I
refer you to Part 3 of this series
for more details.
Fig. 5 shows the pin-out
designations for the external
interfacing board. These match
the mainframe wiring done in
Part 2 of this series.
In order to insure that
builders of the KIM System ful-
ly understand how to utilize the
joystick interface, I have includ-
ed a short program in Fig. 5 that
will enable you to draw with the
joystick on the TV screen. The
sense switches control the col-
ors of the two-color dot that is
moved by the joystick.
This program serves a useful
function as an educational en-
deavor, and that’s about all.
However, I do feel that builders
of the KIM System will find it
useful as a reference. I have
tried to functionally describe
each section with comments.
This completes the hardware
series on my KIM-1 system,
which now contains 17K of
RAM and supports both BASIC
and FOCAL. I’m also in the pro-
cess of implementing a new
language with an ease of use
and a speed of operation some-
where between assembly lan-
guage and BASIC.
Because my system is to be
multilingual, I have chosen to
avoid ROM in favor of RAM for
all functions except the KIM
monitor. I’m also planning sev-
eral surprises that I hope to
share iri the future.*
Address
00 00
02
04
07
09
0C
OF
10
11
12
13
15
18
1A
1C
IE
21
23
25
26
28
2A
2C
2F
31
33
Contents
Label
Mnemonic
LOC PAGE DATA STORE POINTER
:Set mode and starting address for the dazzler
A9 10 INIT LDA #$10
8D OF 80 STA MODE
A9 90 LDA #$90
8D 0E 80 STA BEGADDR
:Get horizontal joystick position
AD 11 80 START LDA JOYHOR
: Place 4 MSB into 4 LSB and save
4A LSR
4A LSR
4A LSR
4A LSR
85 00 STA LOC
:Get vertical joystick position
AD 10 80 LDA JOYVER
: Check for and set up proper page of screen
display
TOP
30 07
A0 20
84 01
4C 25 00
A0 21
84 01
BOTTOM
BMI BOTTOM
LDY #$20
STY PAGE
JMP CONT
LDY #$21
STY PAGE
: Remove MSB and keep only the next four
0A CONT ASL
29 F0 AND #$F0
: Combine LSB and MSB into one word and save
05 00 • ORA LOC
85 00 STA LOC
:Put color (sense switches) into accumulator
AD 80 80 LDA SENSE
:Prepare for an indirect store using 00 and
01 as pointer
A2 00 LDX #$00
: Store color
81 00 STA LOC PAGE
: Begin Again
4C 0C 00 JMP START
Fig. 6. Sample program for drawing on TV using joystick.
NorthStar Software
Maillist
Maillist is a general purpose mailing label program capable of
producing formatted lists for tractor-fed or Xerox type labels.
Maillist will also sort lists for any field.
Price $39.95 on diskette with manual /stock to 14 day delivery.
In-out driver
Dos in-out driver is designed to set up mapped memory video
boards in conjunction with hard copy device. The user may
switch output under software control. Any file directory may be
listed while in BASIC without jumping to dos. Spacebar will stop
output for line by line listings. Designed for use with 3P-t-S and
any tv board.
Price $12.95 on diskette with manual /stock to 14 day delivery.
Register
Register is a cash register and inventory control program. The
software will control a point of sale terminal and printer. It will
search inventory for an item, price and ticket it. Register has
provisions for min-max, automatic reorder, and critical list.
Price $299.95 on diskette with manual
All prices are FOB Santa Barbara, California.
Terms COD Residents add 6 % sales tax and $1.00 shipping.
Alpha Data Systems A48
Box 267, Santa Barbara, Ca. 93102 ■ 805/682-5693
Datapoint 2200 Computer
Full-Assembled
Operating
$ 995 .
#2200 VI used
• Add $25 packing
• Shipped FOB Washington, D.C. Terms check, MO or
charge.
• Guaranteed operating • Program package $49.50
• 8 K memory • Upper and lower case ASCII Display
• Parallel I/O • 80 col. screen • dual tape cassette
drives. The Datapoint 2200 VI is a complete self-con-
tained general-purpose computer. May be used as an
intelligent terminal. While they last, with each 2200 we
will ship, at no additional cost, a printer which may be
adapted to run with the processor (no guarantee on
printer).
ALSO AVAILABLE: Datapoint 3000
CRTs $675.00
TELECOMMUNICATIONS SERVICES CO.
Box 4117, Alexandria , Va . 22303 -rofi
703-683-4019 /TLX 89-623
RAINBOW COMPUTING INC.
Supplier of
Apple
Wave Mate
The Digital Group
Southwest Technical Products
Digital Equipment Corporation
Computer Products
Peripherals and Supplies from
PerSci Computer Devices
Contronix Lear-Siegler
Diablo Multi-Tech
Maxell Texas Instruments
'Scotch' Brand Magnetic Media
Specialists in Design, Implementation and
Support of Custom Hardware/Software
for Business, Educational, and
Personal Use
Consu 1 1 ing/Con tract/Progra mm i ng
Operating Systems/Applications Software
Experts in most major computer
software including
CDC, IBM, PDP
BASIC. COBOL. FORTRAN, PL1
Lisp, Simula, Snobol, SPSS, BMD's
COMPASS, MACRO, 6800, & Z80 assembly languages
10723 White Oak Ave., Granada Hills, Ca. 91344
(213)360-2171 Rin
63
Sheila Clarke
CyberGrafix
518 N. Brand
Glendale CA 91203
What’s Happening
with the IBM Selectric?
Micro Computer Devices has the answer
A rt Childs needed a printer
for a long time. An IBM
Selectric had been Art’s and my
choice for a couple of years. It
was ideal because of its small
size and beautiful print quality.
However, we both were skepti-
cal about printers available for
use with a computer. In most
cases, either the typewriter
was used and reconditioned or
a lot of interface kit assembly
was required. Like many com-
puter-users, Art can’t afford to
risk having an unreliable unit
requiring continual mainte-
nance; nor does he have time to
assemble a kit.
When we first heard about
the SELECTERM, made by
Micro Computer Devices
(MCD), we were impressed that
someone had finally converted
a brand new typewriter for
microcomputers. Because both
IBM and Micro Computer De-
vices provide warranties for
their respective portions of the
device, we decided to obtain a
SELECTERM.
Art’s system consists of an
Altair 8800, dual ICOM floppy-
disk drive, and an ADM CRT ter-
minal. He uses the 3P + S inter-
face board from Processor
Technology. After spending
two hours struggling to deci-
pher the board’s schematics,
which seemed to be written
solely for hardware types, Art
finally called his engineer
friend, Steve Griffis, who came
over and had everything run-
ning in five minutes. Although
the SELECTERM will interface
to any microcomputer, what if
you can’t read the interface
board schematics? Micro Com-
puter Devices is providing a
solution with specific connect-
ing instructions for each inter-
face available on every com-
puter now being sold.
Art’s reaction to the printer
was positive from the moment
the two large cartons were
delivered. One carton held the
Selectric and the other con-
tained the electronics package.
He was impressed with the
packing, which held the units
solidly with formed foam to pre-
vent damage caused in ship-
ping. Opening the flap of the
carton, Art uncovered a sheet
that said STOP, with complete
unpacking and typewriter as-
sembly instructions. Art, in too
big a hurry, merely made a men-
tal note that instructions were
there and, consequently, ran in-
to a little trouble securing the
cover latches of the typewriter.
(Sometimes I wonder if anyone
reads anything before making
panic calls to the man-
ufacturer.)
He was also impressed with
the documentation and the
SELECTERM’s acceptance of
ASCII. With no conversion nec-
essary, Art began writing a driv-
er. It took him five minutes, us-
ing assembly language for
FDOS-lll. He said the only dif-
ference between this printer
and another line printer driver
or hard-copy output driver is
that you might have to put out
some nulls after tabs and line
feed. But it was simple for him
to write the nulls into the driver.
Fig. 1 shows the driver for the^
8080 and 3P + S.
Art commented: “The IBM
print quality is nice. And I like
the fact that I can change type
fonts. Putting the whole thing
together— removing it from the
cartons to putting the cover on
the typewriter and hooking up
the cables— was a half-hour
task. The fact that it requires
64
If we lift the typewriter up off the baseplate , we see the elec-
tronics added to convert the typewriter to a printer.
one parallel port makes it easy.
If you have only one serial port,
which is often the case, you’ll
usually lose it to your print de-
vice. Writing the driver and in-
tegrating it into the software
completed the process. All in
all, it was very easy; every-
thing’s been done for you. The
unit runs very cool, the elec-
tronics box is barely warm to
the touch after running con-
sistently for about three hours,
and it runs cooler than the
typewriter itself.”
I’m using the SELECTERM to
prepare this article for Kilo-
baud ; I am inputting the text in
the computer, from first draft to
the final, edited version. It’s a
pleasure to know I don’t have to
retype this thing two or three
times before I get it right. The
advantages of the SELECTERM
are only evident when I begin to
use it. For example, the sales
literature doesn’t tell me how to
input uppercase and lowercase
letters with a terminal that has
only uppercase. So MCD owner
Shelly Howard pointed out the
ADM has switches beneath the
nameplate. Setting the LC EN
switch enables me to input up-
percase and lowercase for
printer output. I did discover,
however, that the switch must
always be returned to the UC
position after using the Insert
mode of the text editor. After
that little switch is flipped, the
CRT may only see uppercase
characters, but when I hit shift
for uppercase characters, the
printer outputs caps where
they should be— just like using
a typewriter.
After using the SELECTERM
for a couple of weeks, Art and I
ran into difficulty getting clear
print— then it jammed. The
problem was a loose motor
mount. Because the typewriter
portion was under warranty,
IBM service came out and fixed
it at no charge.
How It All Began
To find out how his product
came about I spent some time
talking with Shelly Howard.
Like many other small-scale
manufacturers, Shelly knew
relatively little about micro-
computers two and a half years
ago. In fact, he was preparing
his thesis for his PhD on an IBM
Selectric. After gathering suffi-
cient research data, he wanted
it compiled through a computer
and output on a Selectric that
matched the type of his own
typewriter. He was told by two
computer outfits that IBM had
discontinued making its I/O
device. He was forced to either
scrap his original plans or buy
his own computer. Assuming
the cost of ownership would be
prohibitive, he searched and
discovered the world of
microcomputers. He also
discovered Don Lancaster’s TV
Typewriter Cookbook.
Now They Tell Me!
Although he followed the
book’s instructions to the let-
ter, Shelly failed to get a unit up
and running. He later discov-
ered the book had been based
on theory only; no one in Lan-
caster’s organization had ac-
tually put the theory to prac-
tice. By now Shelly was too
committed to back out, so he
decided to start over with the
help of two design engineers,
Steve Garner and Jimmy Carter
(no, another one).
Months of design develop-
ment, field testing and improve-
ments resulted in production of
a printer with all parts— the
baseplate, actuators, coils,
transformer and linkages —
manufactured by MCD. Finally,
the design was approved by
IBM. That’s why IBM service
will come and fix your printer if
anything goes wrong; you can
also buy yearly service
agreements from IBM after the
warranty expires. For this
reason, MCD will not sell the
SELECTERM in kit form. IBM
has only approved the factory
assembled and tested model.
In Full Swing
First shipments of the
SELECTERM were made in
August 1977; currently about
three per day are delivered to
dealers. The target is five per
day, but the cash-flow situation
is tough with MCD in a contin-
ual fiscal squeeze. Though IBM
sanctioned the design, MCD is
treated like any other individual
consumer, as far as open credit
goes. When you buy in quanti-
ty, with no quantity discount, at
the same price I paid for my
1
0000
J
ROUTINE TO DRIVE SELECTERM WITH 8080 AND 3P + S
2
0000
y
3
0000
DB04
LO:
IN
4
GET STATUS
4
0002
E601
ANI
1
MASK
5
0004
CA0000
JZ
LO
NOT READY
6
0007
79
MOV
A,C
GET CHAR
7
0008
D306
OUT
6
OUTPUT
8
000A
FE09
CPI
9
WAS IT A TAB?
9
oooc
CA1C00
JZ
LOTAB
YES
10
000F
FE08
CPI
8
NO - BACKSPACE?
11
0011
CA1700
JZ
LOLF
YES
12
0014
FE0A
CPI
0AH
NO - LINE FEED?
13
0016
CO
RNZ
NO - RETURN
14
0017
j
15
0017
0E00
LOLF:
MVI
C,0
OUTPUT A NULL
16
0019
C30000
JMP
LO
AND RETURN
17
001C
y
18
001C
C5
LOTAB:
PUSH
B
19
001 D
010004
LXI
B,400H
4 NULLS
20
0020
CD0000
LOTB1:
CALL
LO
OUTPUT
21
0023
05
DCR
B
LAST ONE?
22
0024
C22000
JNZ
LOTB1
NO
23
0027
Cl
POP
B
YES - RESTORE B
24
0028
C9
RET
AND RETURN
25
0029
y
26
0029
0000
END
TOTAL ERRORS = 00
Fig. 1. ICOM 8080/Z-80 Reloc-Macro Assembler Ver. 1.0.
65
1st Row - Uppercase :
1st Row - Lowercase :
)_ +
1234567890- =
2nd Row - Uppercase :
2nd Row - Lowercase :
3rd Row - Uppercase:
3rd Row - Lowercase:
4th Row - Uppercase :
4th Row - Lowercase :
QWERTYUIOPi
qwertyuiopi
ASDFGHJKL : "
asdfghjkl ; '
ZXCVBNM, .?
zxcvbnm, . /
Fig. 2. ASCII character set for SELECTERM output device.
Selectric II, a lot of bucks are
going out the door at one time.
To handle the dilemma, MCD
sells through dealers only, on a
COD basis. Because requests
have been made by some
manufacturers, the firm wants
to produce OEM versions to
specification. Shelly will proba-
bly find investors, or perhaps
release MCD for acquisition by
another company. But he loves
what he’s doing: selling and de-
livering SELECTERMs to deal-
ers across the country.
Competition
Presently, only one other
company in the country sells an
IBM Selectric printer with ASCII
encoding. Other companies of-
fer used Selectrics complete
with interfacing. Even recondi-
tioned units will not qualify for
the IBM Service Agreement.
If you’re looking for a good
printer, this could be it. But
take heed that 15 characters
per second may not be fast
enough. Long listings could
take hours. For most home
computerists, however, speed
may not be a determining fac-
tor in making a printer selec-
tion. And the benefits are
numerous: All the basic
features of the printer include
the special typing element, tab
command, back space, vertical
tab, bell, serial and parallel in-
terfacing, cable sets and soft-
ware in PROM within the elec-
tronics. Also included is a
special ASCII typing element
that IBM has produced to MCD
specifications. Fig. 2 shows an
output of the character set.
The price of $1750 appears
prohibitive, until you consider
that you’ll be using an extreme-
ly well-designed unit that will
last for years— type fonts are
changed at will, no special
paper is needed, IBM ribbon is
easy to order, and service is vir-
tually hassle-free.
Options
The same extras as those of-
fered by IBM, including dual
When the typewriter cover is off, the SELECTERM looks about like
another Selectric II. Here it sits alongside Art’s Altair 8800 with
cabling interface to the 3P + S.
pitch and correcting feature,
can be ordered for your SELEC-
TERM. MCD has developed a
noise-reduction feature (recom-
mended if you live in a residen-
tial neignborhood).
Tractor-feed platen and
RS-232 interface are also being
offered as options.
After using the SELECTERM
a great deal for two months
now, Art and I are definitely
convinced that we did a good
thing for his computer. And a
nice plus is that we now have a
second typewriter— that is,
when it’s not being used with
the computer.*
§06 § §
§ § §@ §
§ 0000
§ § § 00
000 0 §
§00000
000 § 0
0 00 0 0
§ 0 0 0 00©
0 0 00 §
0©0 © ©
NATIONWIDE CLASSIFIED AD NEWSLETTER
MAILED 1U CLASS EVERY THREE WEEKS
* ARE YOU LOOKING FOR? *
* - LOW COST USED COMPUTER EQUIPMENT *
* - " " NEW *
* - ADVANCED INFORMATION ON NEW PRODUCTS *
* - NEW SOFTWARE: UTILITIES/BUSINESS/etc. *
* - A WAY TO ASK FOR INFORMATION OR AID *
* - INFORMATION ON COMPUTER CLUB MEETINGS *
* *
* IT'S ALL IN ON .LINE! *
* *
* 18 ISSUES (1 Yr. )-$3.75 36 ISS.-$7.00 *
* Sample on 4eque &t, on addzd to tubAcsvip. *
* 2 4 6 9 5 SANTA CRUZ HWY. *
* LOS GATOS, CA 95030 *
* 02
APPLE
OWNERS
Interactive Trap and Chase games for
two people. Software allows choice of
speed and points and requires 8K RAM.
Hardware consists of two game control
boxes. Each box has four micro switches
and specially designed printed circuit
with 5 feet of cable completely assembled.
Plugs right into your game I/O connector.
Control boxes can also be used with the
game Dragon Maze
Hardware and software cassette —
$49.95. (Texas residents add 5% sales
tax.) 90 day warranty parts and labor.
Guaranteed 30 day delivery. Allow time
for processing personal checks. Send
check or money order to:
B & G Interfaces
P.O. Box 59364
Northhaven Sta.
Dallas, Tx. 75229
PROFESSIONAL QUALITY
AUDIO CASSETTES
-^FOR TARBELL , DGS ,KC ,ETC
DON’T WASTE YOUR MONEY
ON CHEAP TAPES,
o EXTENSIVE TESTING HAS
RESULTED IN SELECTION
■OF THIS TAPE FOR BAUD
RATES IN EXCESS OF 1200
SUPER HIGH DENSITY
HIGH FREQUENCY RESPONSE
LOW NOISE. SIGNAL TO
NOISE RATIO 54 DB
-TENSILIZED POLYESTER
"BASE .69 MILS THICK
SONIC WELDED CASSETTE
O 30 MIN TAPES 3 AT $5.50
10-$15.50,25-$35.00 PPD
PITTS ENTERPRISES
1516K BOWEN ST.
LONGMONT, CO. 80501
P24
66
TIMESHARING
The 80-1 03A works both ways. Your system can call a timesharing service and communicate as an intelligent
terminal OR your S-100 system can be the timesharing system where the 80-103A answers the phone and
communicates with terminals or other processors.
80-103A DATA COMMUNICATIONS ADAPTER
The 80-103A DATA COMMUNICATIONS ADAPTER was developed to function as an S-100 bus
compatible serial interface incorporating a fully programmable modem and Telco interface. These
functions are usually accomplished by the use of two separate modules: 1) a serial I/O board, and
2 ) an external modem. By combining these features on a single board, the 80-103A can offer
microcomputer applications significant cost/performance advantages over other implementations.
• FULLY PROGRAMMABLE FEATURES
• AUTOMATED DIALING AND AN-
SWER
• ORIGINATE OR ANSWER MODE
• 1 10-300 BIT/SEC DATA RATES
• CHARACTER FORMAT AND PARITY
• ERROR DETECTION
• FULLY BUFFERED, OUTPUTS DRIVE
25 S-100 BUS LOADS
• STANDARD U.S. FREQUENCIES
• FULL TELCO COMPATIBILITY WHEN
ATTACHED TO DAA
• COMPATIBLE WITH EXISTING TELE-
TYPES AND TIME SHARING MODEMS
• ALL DIGITAL MODULATION AND
DEMODULATION. NO ADJUSTMENTS
REQUIRED.
DC Hayes Associates offers a full range of capabilities
for solving your information handling problems.
Whether your problem is large or small, we will apply
innovative techniques for finding the best solution.
Contact us about our products and services.
C Hayes
PRICES:
Assembled 80-103A with 48 hour burn in
and 90 day warranty is $279.95
Bare Board with manual is $49.95
H20
P.0. Box 9884 • Atlanta, Georgia 30319 • (404) 231-0574
67
The Top-Down
Approach
with some practical examples
Dr. Lance A. Leventhal
PO Box 1258
Rancho Santa Fe CA 92067
I n Kilobaud No. 14 (“Why
Structured Programming?” p.
84), I discussed structured pro-
gramming, a method for mak-
ing the logic of large programs
simple and repetitive, thereby
making them easier to debug
and test. But a further problem
in writing large programs is
how to put sections of the pro-
grams together. This article
describes a widely used
method called top-down
design, by which the program-
mer starts with an overall
outline of the program and pro-
ceeds to steadily describe each
section in greater detail, debug-
ging and testing along the way
in an integrated manner.
Modular Programming
Obviously, a large program
can only be written by dividing
it into sections. No one (I hope)
would simply write the er*ire
program and then see if it
worked. Clearly, a better idea is
to write a small section, see if
that works, correct it, write
another small section, and so
on. This procedure is known as
modular programming and the
sections of the program are
called modules.
Some typical modules in an
overall accounting, game,
word-processing or instruc-
tional program might be: I/O
routines, file-handling routines,
mathematical calculations,
string-handling routines, table
searches, sorting routines,
table lookup and list pro-
cessing.
The advantages of modular
programming are clear.
1. You can check the
modules individually and be
sure they work properly. Thus,
you can assume that any errors
in the overall program are in the
connections or the supervisor
program.
2. You can build a library of
modules that will be useful in
other p r ograms. Many of the
previously mentioned modules
will be needed frequently.
3. You can use modules that
you have previously developed,
found in books or magazine ar-
ticles, or borrowed from
friends. You can also use mod-
ules such as file handlers, code
converters and I/O handlers
that comprise part of your mon-
itor or operating system.
4. You can plan program
development and have a
reasonable idea of how much
progress you have made and
what the major stumbling
blocks are.
5. You can eliminate many
simple errors at an early stage.
Modular programming has
serious disadvantages, though.
Somehow, the modules never
quite seem to fit together at the
end. Different modules may use
different registers, memory
locations or subroutines. Some
may wipe out results that
others need or not use data that
others provide. Module integra-
tion often turns out to be a big
task you must struggle with
after everything seems to be
done.
The problem of integrating
modules is independent of the
problem of testing and debug-
ging them. The modules may all
work separately, but still not
work together. The catch is that
the original debugging and
testing checks the workings of
the module out of context (i.e.,
all by itself rather than as part
of a complete program).
In fact, debugging and
testing a module in isolation
can be quite difficult. A game
program, for example, may con-
sist of the following modules:
(1) determine initial conditions,
(2) read and check proposed
move (see if it is valid), (3) deter-
mine new conditions, (4) print
status.
But how can you write the
routine that reads and checks
the proposed move unless you
know the previous state of the
game and can see the new
state? How will you be able to
tell if the MOVE module is work-
ing properly? Typically, you will
have to either manually enter
the required data and examine
the results or write special pro-
grams to perform those tasks.
These special programs (some-
times called driver programs)
can save a lot of manual effort;
however, they introduce extra
work and may act quite dif-
ferently from the real routines
for which they substitute. (Note
that you don’t save the driver
programs; you throw them
away when the job is done.)
Clearly, the problem of com-
bining modules is even more
serious in large commercial
programming projects. Not on-
ly can the number of modules in
a project be very large, but also
many programmers may be in-
volved in writing them. Now the
problem is to integrate mod-
ules written by people with dif-
ferent styles, different levels of
expertise, different docu-
mentation methods and dif-
ferent interpretations of tasks.
Top-down Design
Most commercial program-
ming shops now use some ver-
sion of top-down design. This
method differs from the more
traditional bottom-up design
(see Fig. 1) in which the specific
modules are written before
they are integrated into more
complex programs. Top-down
design (see Fig. 2) proceeds as
follows:
1. The overall supervisor pro-
gram is written, debugged and
tested. Major subprograms are
replaced by program stubs that
may produce the answer to a
selected problem, record the
entry or do nothing at all.
2. Each stub is then similarly
68
(I) WRITE THE MODULES
(I) WRITE THE OVERALL PROGRAM
(2) COMBINE THEM INTO PROGRAMS OF INCREASING COMPLEXITY (2) EXPAND EACH SUBPROGRAM, ADDING DETAIL AS REQUIRED UNTIL THE
PROBLEM DEFINITION IS MET.
Fig. 1. The procedure for bottom-up design.
Fig. 2. The procedure for top-down design.
expanded, with debugging and
testing occurring at each step.
Advantages of Top-down
Design
The advantages of top-down
design are:
• It modularizes debugging,
testing and integration, as
well as coding (the writing of
instructions).
• It allows subprograms to be
debugged and tested in the
actual environment of the
entire program. No special
debugging and testing
programs (or drivers) are
needed to provide data or to
interpret results.
• It results in overall program
logic being checked first.
This often means that the
programmer can immediate-
ly discover and eliminate in-
consistencies and miscon-
ceptions that otherwise may
be very difficult to find and
correct (after all the
modules have been written).
• It provides a systematic
framework for program
development and testing. It
gives the programmer a firm
idea of how much of the task
has been accomplished.
Disadvantages of Top-down
Design
Of course, like all methods,
top-down design has disadvan-
tages. Among these are:
• A suitable program stub
may be difficult to write, par-
ticularly if it must appear in
many different places and
produce many different inci-
dental effects.
• The top-down expansion
may not mesh well with
hardware or already existing
software.
• Errors in the overall program
can have catastrophic ef-
fects on the entire project.
Often critical design deci-
sions must be made early
before you know what prob-
lems exist (or will be
created) at the lower levels.
Furthermore, top-down
design assumes a simple pro-
gram structure with indepen-
dent subsections (i.e., a tree
structure, as shown in Figs. 1
and 2). Some programs
(perhaps even most) can logi-
cally be constructed in that
manner. But there is no proof
that all, or even most, programs
can be. Often programs have in-
terconnections at all levels that
defy simple analysis.
Of course, top-down design
is no panacea; it provides
neither rules nor guidance for:
(1) dividing programs into
modules that can be written in-
dependently of other modules;
(2) writing the modules (here,
structured programming
comes into play); (3) defining or
using data structures ... in
many situations, the structure
of the data may be more impor-
tant and more difficult to deter-
mine than the structure of the
program.
But top-down design does
provide a systematic frame-
work, rather than a haphazard
approach. This framework has
been shown to significantly in-
crease programmer productivi-
ty in the commercial world. Fur-
thermore, it seems to result in
programs that have clearer
logic and are easier to test,
debug, extend and use. Of
course, programmers should
never disdain a little bottom-up
design where that method per-
mits better utilization of hard-
ware, existing software or other
resources. The aim of program-
ming is to produce programs that
work, not to follow the tenets of
one methodology or another.
Much of what we have said
so far about top-down design is
vague. Now let us see how it
works in a real example.
The Vote Analysis Program
The purpose of this program
is to count ballots and print the
totals in decreasing order-
starting with the candidates
who received the most votes. C
is the number of candidates,
and the ballots are coded as
follows:
0— a blank ballot (no vote for
any candidate).
1 to C— vote for the indicated
candidate.
C + 1— vote for a write-in
candidate.
C + 2— illegal vote (two or
more candidates marked).
C + 3— special marking for
last (dummy) ballot.
Fig. 3 shows the initial pro-
gram flowchart. The important
variables are: N (I)— number of
votes for candidate I, V— total
number of votes, M (I)— can-
didate numbers for rank-
ordering.
We have not tried here to
make the programs particularly
efficient or to make the I/O
realistic. Rather, we have tried
to show how program develop-
ment proceeds, starting with an
overall skeleton program and
continuing through ever-in-
creasing levels of detail. The
language is a simple version of
BASIC that should run on most
computers.
Initial Program
Fig. 4 contains the initial pro-
gram listing. The three major
sections of the program-
counting, ordering and out-
put— have been replaced by
program stubs that simply
mark those sections that have
been entered. We can test the
overall program logic by enter-
ing a value for the number of
candidates, C, and running the
^ START ^
ALL TOTALS
s 0 NUMBER
OF VOTES
(N)>0
READ
NUMBER OF
CANDIDATES
COUNT VOTES
OROER
( TOTALS
CEO
Fig. 3. Initial flowchart for the
vote-analysis program.
69
LIST
10
DIM N ( 20 ) # M ( 20 )
15
REM NUMBER OF VOTES (V) = 0
20
LET V= 0
25
REM GET NUMBER OF CANDIDATES (C)
30
PRINT "NUMBER OF CANDIDATES = "J
35
INPUT C
43
REM CLEAR ALL VOTE COUNTERS
45
FOR 1= 1 10 C* 2
50
LET N ( I > = 0
55
NEXT I
60
REM COUNT VOTES
65
GOSUB 1000
70
REM ORDER VOTE TOTALS
75
GOSUB 2000
80
REM OUTPUT TOTALS
85
GOSUB 3000
999
END
1000
REM VOTE COUNTING PROGRAM
1010
PRINT "ATTEMPTED VOTE COUNTING"
1020
RETURN
20 00
REM TOTAL ORDERING PROGRAM
2010
PRINT "ATTEMPTED ORDERING"
20 20
RETURN
3000
REM OUTPUT ROUTINE
30 10
PRINT "REACHED OUTPUT ROUTINE"
3020
RETURN
9999
END
RUN
NUMBER OF CANDIDATES = ? 0
ATTEMPTED VOTE COUNTING
ATTEMPTED ORDERING
REACHED OUTPUT ROUTINE
READY
Fig. 4. Initial listing for the vote-analysis program. All the sub-
programs are left as unexpanded stubs.
program (note the RUN results
at the bottom). In fact, there
was a slight error Initially
caused by the omission of the
final END statement. This error
was quickly corrected before
any stubs were expanded.
The First Level of Expansion
Fig. 5 is the flowchart of the
expanded vote-counting pro-
^ START ^
INPUT A VOTE
N (VOTE)*
N (VOTEJ+ I
Fig. 5. Flowchart for the vote-
counting subprogram.
gram. Here there are three
cases to consider:
1. The last ballot (marked
with the number C + 3) is not
counted in the totals.
2. Blank ballots (marked by
zero) are included in the total
number of votes but are not
credited to any category.
3. Other ballots must be
credited to the appropriate
category (i.e., to a candidate,
write-in category or improperly
marked category).
Fig. 6 contains the BASIC
program with the vote-counting
stub expanded. We checked
this program with the data in
Example 1 (see the results at
the bottom of Fig. 6).
Fig. 7 contains the BASIC
program with the output stub
expanded. This program was
also checked with cases 1 and
2. Note the added statement
3020 IF C = 0 THEN 3045
This correction means that if
there are no candidates, the
program does not print head-
ings, a list of candidates or vote
totals. Note that the case
LIST
10 DIM N ( 20 > > M ( 20 )
IS HEM NUMBER 0F VOTES CV) = 0
20 LET V= 0
25 REM GET NUMBER 0F CANDIDATES <C>
30 PRINT "N UMR ER 0E CANDIDATES = "I
35 INPUT C
43 REM CLEAR ALL VOTE COUNTERS
45 FOR 1= I TO C* 2
50 LET N< I ) = 0
55 NEXT I
60 REM COUNT VOTES
65 G0SUB 1000
70 REM ORDER VOTE TOTALS
75 G0SUB 2000
80 REM OUTPUT TOTALS
85 GOSUB 3000
999 END
1000 REM VOTE COUNTING PROGRAM
1005 REM FETCH NEXT VOTE (J>
1010 PRINT "NEXT VOTE IS")
101 5 INPUT J
1020 REM DONE IF VOTE IS ENDING MARK <C*3)
1025 IF J = C* 3 THEN 1065
1030 REM ADD VOTE TO TOTAL CV)
1035 LET V=V* 1
10 40 REM IGNORE VOTE IF BALLOT UN MARK D (J = 0>
1045 IF J= 0 THEN 1010
1050 REM ADD VOTE TO APPROPRIATE TOTAL
1055 LET N C J ) =N ( J ) ♦ 1
1060 GOTO 1010
1065 RETURN
2000 REM TOTAL ORDERING PROGRAM
2010 PRINT "ATTEMPTED ORDERING"
2020 RETURN
30 0 0 REM OUTPUT ROUTINE
3010 PRINT "REACHED OUTPUT ROUTINE"
3020 RETURN
9999 END
RUN
NUMBER OF CANDIDATES = ? 0
NEXT VOTE IS? 3
ATTEMPTED ORDERING
REACHED OUTPUT ROUTINE
READY
RUN
NUMBER OF CANDIDATES = ? 1
NEXT VOTE IS? 1
NEXT VOTE IS? 4
ATTEMPTED ORDERING
REACHED OUTPUT ROUTINE
REA DY
Fig. 6. Listing for the vote-analysis program with the vote-
counting subprogram expanded.
without a candidate, although
it seems useless, is by no
means an uncommon situation
in real elections, particularly at
the local level. The results from
this expanded program are in
Fig. 8.
Fig. 9 is a flowchart for the
first expansion of the rank-
ordering routine. The idea is to
keep interchanging pairs of
elements until all pairs are in
the correct order (i.e., largest
number first). Flag F is cleared
initially and set to 1 if an inter-
change is performed. So, if F =
1 at the end of a pass through
the list, another pass is
necessary. If F = 0 at the end,
the list must be in order.
Although this may appear an
unsophisticated sorting meth-
od, it is perfectly acceptable for
short lists like the ones han-
dled by this program. The
number of candidates in an
election rarely exceeds ten.
Note that no sorting is
necessary if there is only one
candidate or are none.
Fig. 10 is the BASIC program
with the ordering routine ex-
70
RUN
NUMBER 0F
CANDIDATES = ? 0
NEXT vote:
IS? 3
ATTEMPTED
ORDERING
NUMBER 0F
CANDIDATES = 0
NUMBER 0F
VOTES = 0
NUMBER 0F
WRITE-INS = 0
NUMBER 0F
READY
IMPROPER BALLOTS =
0
RUN
NUMBER 0F
CANDIDATES = ? 1
NEXT VOTE
IS? 1
NEXT VOTE
IS? A
ATTEMPTED
ORDERING
NUMBER 0F
CANDIDATES = 1
NUMBER 0F
VOTES = 1
CANDIDATE
NUMBER VOTE TOTAL
1
1
NUMBER 0 F
WRITE-INS = 0
NUMBER 0F
READY
IMPROPER BALLOTS =
0
RUN
NUMBER 0F
CANDIDATES = ? 2
NEXT V0TE
IS? \
NEXT V0TE
IS? 1
NEXT V0TE
IS? 2
NEXT VOTE
IS? 5
ATTEMPTED
ORDERING
NUMBER OF
CANDIDATES - 2
NUMBER 0F
VOTES - 3
CAN DI DATE
NUMBER VOTE TOTAL
1
2
2
1
NUMBER 0F
WRITE-INS = 0
NUMBER OF
READY
IMPR0PFR BALLOTS =
0
Fig. 8. Results from the program of Fig. 7.
LIST
10
DIM N ( 20 ) # MC20 )
15
REM NUMBER OF VOTES <V> = 0
90
LET V= 0
25
REM GET NUMBER 0F CANDIDATES <C>
30
PRINT "NUMBER 0F CANDIDATES = "1
35
INPUT C
O
REM CLEAR ALL V0TE C0UNTERS
45
FOR 1= 1 T0 C» 2
50
LET N ( I ) = 0
55
NEXT I
60
REM C0UNT VOTES
65
G0SUB 1000
70
REM ORDER VOTE TOTALS
75
G0SUB 2000
TO
REM OUTPUT TOTALS
85
G0SUB 3000
999
END
1000
REM V0TF COUNTING PROGRAM
100 5
REM FETCH NEXT VOTE ( J )
1010
PRINT "NEXT VOTE IS"I
101 5
INPUT J
1020
REM DONE IF VOTE IS ENDING MARK (C*3>
1025
IF J = C* 3 THEN 1065
10 30
REM ADD VOTE TO TOTAL CV>
1035
LET V= V-*- 1
10 40
REM IGNORE VOTE IF BALLOT UNMARKED <J=
0)
1045
IF J= 0 THEN 1010
10 50
REM ADD VOTE TO APPROPRIATE TOTAL
10 55
LET N(J)=N(J)+ 1
1060
GOTO 1010
1065
RETURN
2000
REM TOTAL ORDERING PROGRAM
2010
PRINT "ATTEMPTED ORDERING"
20 20
RETURN
3000
REM OUTPUT ROUTINE
300 5
PRINT "NUMBER OF CANDIDATES = "1C
3010
PRINT "NUMBER OF VOTES * "IV
301 5
REM SKIP CANDIDATE TOTALS IF NO CANDIDATES
3020
IF C s 0 THEN 30 45
302 5
PRINT "CANDIDATE NUMBER VOTE TOTAL"
3030
FOR 1= 1 TO C
30 3 5
PRINT TAB ( 5)# I# TAB( 25>#NCI>
30 40
NEXT I
30 45
PRINT "NUMBER OF WRITE-INS = "JN<C* 1>
30 50
PRINT "NUMBER OF IMPROPER BALLOTS * "INCC* 2)
30 5 5
RETURN
99 9 9
END
Fig . 7.
Listing for the vote-analysis program with
the vote-
counting and output subprograms expanded.
CASE 1. NO CANDIDATES, NO VOTES
C = 0
V = 3 (ENDING MARKER)
CASE 2. ONE CANDIDATE, ONE VOTE
C = 0
V = 1
V = 4 (ENDING MARKER)
Example 1.
panded. Note that the inter-
change subroutine is left as a
program stub. It will be ex-
panded later. For some simple
cases for checking this pro-
gram, see Example 2. Fig. 11
shows the results from this pro-
gram. Note that an interchange
was attempted in Case 4, but
not in Case 3.
The Second Level of Expansion
Fig. 12 shows the program
with the interchange stub ex-
panded. Statement 3035 now
prints the identification
number M(l), which is inter-
changed, but statements 2010
and 2033 had to be changed to
give a value to M(l) when there
is only one candidate.
Fig. 12 also contains a fur-
ther expansion of the ordering
routine (see flowchart in Fig.
13) to handle more efficiently
the simple, but common, case
where there are only two can-
didates. Further expansions
could check for erroneous
values of number of candidates
Fig. 9. Flowchart for the rank-ordering subprogram.
71
LIST
10 DIM N(20)#M(20>
15 REM NUMBER OF VOTES ( V) = 0
20 LET V= 0
25 REM GET NUMRER 0F CANDIDATES <C>
30 PRINT "NUMBER 0E CANDIDATES = "J
35 INPUT C
43 REM CLEAR ALL VOTE COUNTERS
45 E0R I« I T0 C* 2
SO LET NC I )= 0
55 NEXT I
60 REM COUNT VOTES
65 G0SUB 1000
70 REM ORDER VOTE TOTALS
75 GOSUB 2000
80 REM OUTPUT TOTALS
85 GOSUB 3000
999 END
1000 REM VOTE COUNTING PROGRAM
1005 REM FETCH NEXT VOTE (J>
1010 PRINT "NEXT VOTE IS"J
101 5 INPUT J
1020 REM DONE IF VOTF IS ENDING MARK <C*3> 3010
1025 IF J = C* 3 THEN 1065 X15
1030 RFM ADD VOTE TO TOTAL (V) 3020
1035 LET V=V* 1 302b
1040 REM IGNORE VOTE IF BALLOT UNMARKED <J = 0> 30 30
1045 IF J= 0 THEN 1010 3035
1050 REM ADD VOTE TO APPROPRIATE TOTAL X 40
1055 LET N< J ) =N < J ) ♦ 1 3045
10 60 GOTO 1010 30 50
10 65 RETURN 30 55
aOOO REM TOTAL ORDERING PROGRAM 9999
REM NO ORDERING NECESSARY IF ZERO OR ONE CANDIDATES
IF C< 2 THEN 2085
REM ASSIGN MARKERS TO CANDIDATES FOR SORTING
FOR 1= 1 TO C
L FT M(I)=I
NEXT I
REM SORT VOTE TOTALS
LET F= 0
FOR 1= 1 TO C- 1
REM CHECK IF TOTALS ARE IN ORDER
IF NCI) > = N ( I 1) THEN 2070
REM IF OUT OF ORDER# INTERCHANGE PAIR
GOSUB 2 500
NEXT I
REM DO ANOTHER PASS IF ANY INTERCHANGES OCCURRED
IF F= 1 THEN 20 40
RETURN
REM INTERCHANGE TOTALS# MARKERS FOR ORDERING
PRINT "ATTEMPTED INTERCHANGE"
RETURN
REM OUTPUT ROUTINE
PRINT "NUMBER OF CANDIDATES = "*C
PRINT "NUMBER OF VOTES * "IV
REM SKIP CANDIDATE TOTALS IF NO CANDIDATtS
IF C= 0 THEN 3045
PRINT "CANDIDATE NUMBER VOTE TOTAL"
FOR 1= t TO C
PRINT TAB ( 5) # I # TAB C 25)#N(I)
NEXT I
PRINT "NUMBER OF .W K I I E-INS = "*N<C* 1)
PRINT "NUMBER OF IMPROPER BALLOTS = ")NCC«- 2)
RETURN
END
200 5
2010
201 5
20 20
2025
20 30
20 3 5
20 40
20 4 5
20 50
2055
20 60
20 6 5
2070
2075
20 80
20 8 5
2500
2510
2520
3000
300 5
Fig. 10. Listing of vote-analysis program with all subprograms expanded by one level.
RUN
NUMBER 0F CANDIDATES = ? 0
NEXT V0TE IS? 3
NUMBER 0F CANDIDATES = 0
NUMBER 0F V0TES = 0
NUMBER 0F WRITE-INS = 0
NUMBER 0F IMPR0PFR BALL01S
READY
0
RUN
NUMBER 0F
NEXT V0TE
NEXT V0TE
NUMBER 0F
NUMBER 0F
CANDIDATE
1
NUMBER 0F
NUMBER 0F
READY
CANDIDATES = ? 1
IS? 1
IS? A
CANDIDATES = 1
V0TES = 1
NUMBER V/0TE T0TAL
1
WRITE-INS * 0
IMPR0PER BALL0TS =
0
RUN
NUMBER 0F
NEXT V0TE
NEXT V0TE
NEXT V0TE
NEXT V0TE
ATTEMPTED
NUMRER 0F
NUMBER 0F
CANDIDATE
1
2
NUMBER 0F
NUMBER 0F
READY
CANDIDATES = ? 2
IS? 1
IS? 2
IS? 2
IS? 5
INTERCHANGE
CANDIDATES = 2
V07ES = 3
NUMBER V0TF TOTAL
1
2
WRITE-INS = 0
I MPR0P E'R BALL 0 TS =
0
Fig. 11. Results from the program of Fig. 10.
CASE 3. TWO CANDIDATES. THREE VOTES (2
FOR NUMBER 1, 1 FOR NUMBER 2)
C = 2
V = 1
V = 1
V = 2
V = 5 (ENDING MARKER)
CASE 4. TWO CANDIDATES, THREE VOTES (1
FOR NUMBER 1, 2 FOR NUMBER 2)
C = 2
V = 1
V = 2
V = 2
V = 5 (ENDING MARKER)
Example 2.
(less than zero or more than the
program can handle) and er-
roneous data (values that are
undefined). Other expansions
could check for ties, handle
cases where more than one
vote is allowed (e.g., vote for
four of the above) and identify
the ballots on which write-ins
were marked.
Conclusion
Top-down design is a method
for designing, debugging and
testing large programs. It re-
quires the programmer to start
with the overall program logic
and to continue expanding sub-
programs until the task is fully
defined. Each level is checked
in its actual working environ-
ment before the next level is at-
tempted. Thus, integration of
modules and system-level
debugging and testing are per-
formed throughout program
development rather than all at
the end. Program stubs replace
unexpanded programs or mod-
ules at each level. Top-down
design is a systematic ap-
proach to writing large pro-
grams. Personal computer
users should carefully consider
its use when attempting com-
plex projects. ■
References
1. J. K. Hughes and J. J.
Michtom, A Structured Ap-
72
LIST
20 38
IF C= 2 THEN 2090
10 DIM NC20)#M(20)
20 40
LET F« 0
15 REM NUMBER 0F VOTES < V) = 0
20 4 5
FOR 1= 1 TO C- 1
20 LET V= 0
20 50
REM CHECK IF TOTALS ARE IN ORDER
25 REM GET NUMBER OF CANDIDATES <C>
20 5 5
IF NCI) >*N ( I ♦ 1) THEN 2070
30 PRINT "NUMBER OF CANDIDATES = "J
20 60
REM IF OUT OF ORDER# INTERCHANGE PAIR
35 INPUT C
20 6 5
GOSUB 2 500
40 REM CLEAR ALL VOTE C0UN1ERS
2070
NEXT I
45 FOR 1= 1 T0 C* 2
207 5
REM DO ANOTHER PASS IF ANY INTERCHANGES OCCURRED
bO LET N < I ) = 0
2080
IF F= 1 THEN 20 40
55 NEXT I
20 8 5
RETURN
60 REM COUNT VOTES
20 9 0
REM ORDER TOTALS FOR TWO CANDIDATES ONLY
65 GOSUB 1000
20 9 5
REM NO PROBLEM IF ALREADY IN ORDER
70 REM ORDER VOTE TOTALS
2100
IF N ( 1 )> = NC 2) THEN 2120
75 GOSUB 2000
210 5
REM IF OUT OF ORDER# INTERCHANGE
80 REM OUTPUT TOTALS
21 10
LET 1= 1
85 GOSUB 3000
21 1 5
GOSUB 2500
999 END
2120
RETURN
1000 REM VOTE COUNTING PROGRAM
2500
REM INTERCHANGE TOTALS# MARKERS FOR ORDERING
1005 REM FETCH NEXT VOTE (J)
250 5
REM MARK THAT INTERCHANGE OCCURRED CF=I>
1010 PRINT "NEXT VOTE IS"!
2510
LET F= 1
101 5 INPUT J
251 5
REM INTERCHANGE TOTALS
1020 REM DONE IF VOTE IS ENDING MARK (C*3>
2520
LET T=NCI)
1025 IF J = C+ 3 THEN 1065
252 5
LET N(I>=N(I«- 1 )
1030 REM ADD VOTE TO TOTAL CV)
2530
LET NC I* 1 > = T
103 5 LET V=V+ 1
253 5
REM INTERCHANGE MARKERS
10 40 REM IGNORE VOTE IF BALLOT UNMARKED <J =
0)
2540
LET T*M ( I )
1045 IF J= 0 THEN 1010
2545
LET M< I >=M< !♦ 1 )
10 50 REM ADD VOTE TO APPROPRIATE TOTAL
2550
LET MCI* 1 ) = T
1055 LET N(J)=N(J>* 1
2555
RETURN
1060 GOTO 1010
30 0 0
REM OUTPUT ROUTINE
1065 RETURN
30 0 5
PRINT "NUMBER OF CANDIDATES = "! C
2000 REM TOTAL ORDERING PROGRAM
3010
PRINT "NUMBER OF VOTES = "1 V
2005 REM DONE IF NO CANDIDATES
XI 5
REM SKIP CANDIDATE TOTALS IF NO CANDIDATES
2010 IF C= 0 THEN 2085
X20
IF C= 0 THEN 30 45
2015 REM ASSIGN MARKERS TO CANDIDATES FOR SORTING
3025
PRINT "CANDIDATE NUMBER VOTE TO 1 AL"
2020 FOR 1= 1 TO C
X 30
FOR 1= 1 TO C
2025 LET M( I ) s I
X35
PRINT TAB C 5)#M( I )# TABC 25>#NCI>
20 30 NEXT I
X 40
NEXT I
2031 REM NO ORDERING NECESSARY IF ONLY ONE
CAN D1 DATE
X 4 5
PRINT "NUMBER OF WRITE-INS = "INCC* 1)
20 3 3 IF C* 1 THEN 208 5
X SO
PRINT "NUMBER OF IMPROPER BALLOTS = "!NCC* 2)
20 3 5 REM SORT VOTE TOTALS
X55
RETURN
2036 REM HANDLE CASE OF ONLY TWO CANDIDATES
SLPAKA 1 ELY
9999
END
Fig. 12. Listing of vote-analysis program with improved rank-ordering subprogram. The subprogram now handles the case of two
candidates more efficiently.
Fig. 13. Flowchart of the improved rank-ordering subprogram.
proach to Programming,
Prentice-Hall, Englewood Cliffs
NJ, 1977.
2. E. Yourdon, Techniques of
Program Structure and Design,
Prentice-Hall, Englewood Cliffs
NJ, 1975.
3. B. W. Kernighan and P. J.
Plauger, The Elements of Pro-
gramming Style, McGraw-Hill,
NY, 1974.
4. E. J. Miller, Jr., and G. E.
Lindamood, “Structured Pro-
gramming: Top-down Ap-
proach,” Datamation , Decem-
ber 1973, pp. 55-57.
5. R. W. Ulrickson, “Solve Soft-
ware Problems Step by Step,”
Electronic Design, January 18,
1977, pp. 54-58.
6. L. A. Leventhal, 8080 A/8085
Assembly Language Program-
ming , Osborne and Associates,
Berkeley CA, 1978.
RUN
NUMBER 0F CANDIDATES = ? 2
NEXT VOTE IS? 1
NEXT VO TE IS? 1
NEXT VOTE IS? 2
NEXT VOTE IS? 5
NUMBER OF CANDIDATES = 2
NUMBER OF VOTES = 3
CANDIDATE NUMBER VOTE TOTAL
1 2
2 1
NUMBER OF WRITE-INS = 0
NUMBER OF IMPROPER BALLOTS = 0
REA DY
RUN
NUMBER OF CANDIDATES = ? 2
NEXT VOTE IS? 1
NEXT VOTE IS? 2
NEXT VOTE IS? 2
NEXT VOTE IS? 5
NUMBER OF CANDIDATES = 2
NUMBER OF VOTES = 3
CANDIDATE NUMBER VOTE TOTAL
2 2
1 1
NUMBER OF WRITE-INS = 0
NUMBER OF IMPROPER BALLOTS = 0
REA DY
Fig. 14. Results from the program of Fig. 12.
73
wire wrapping
center
lor quality electronic parts and tools.
HOBBY WRAP
TOOL
Wire-wrapping, stripping, unwrapping tool for
AWG 30 on. 025 (0,63mm) Square Post.
Regular Wrap
WSU-30
$6.95
Modified Wrap
WSU-30M
$7.95
tery SW
WIRE-WRAPPING TOOL
For .025" (0,63mm) sq. post
“MODIFIED”wrap, positive
indexing, anti-overwrapping
device.
Battery w|j
wire
wrapping
tool )
COMPLETE
WITH BIT
AND SLEEVE
For AWG 30
BW-630
$34.95*
For AWG 26-28
BW 2628
$39.95*
Bit for AWG 30
BT-30
$3.95
Bit for AWG 26-28
BT-2628
$7.95
USE "C" SIZE Nl CAD BATTERIES
(NOT INCLUDED)
WIRE-WRAPPING KITS
Contains: Hobby Wrap Tool WSU-30,
(50 ft.) Roll of wire
Prestripped wire 1" to 4"
lengths (50 wires per package)
stripped 1" both ends.
Wire Wrapping Kit. ( Blue)
WK 2 B
$12.95
Wire Wrapping Kit. (Vetlow)
WK 2 Y
$1295
Wire Wrapping Kit. (White)
WK 2 W
$12.95
Wire Wrapping Kit (Red)
WK-2R
$12.95
ROLLS OF WIRE
Wire for wire-wrapping AWG-30
(0.25mm) KYNAR* wire, 50 ft. roll,
silver plated, solid conductor,
easy stripping.
30 AWG Blue Wire. 50ft Roll
R 308-0050
$198
30 AWG Yellow Wire 50ft Roll
R30Y0050
$198
30 AWG White Wire 50ft Roll
R 30W 0050
$198
30 AWG Red Wire. 50ft Roll
R 30R-0050
$198
WIRE-WRAPPING KIT
Contains: Hobby Wrap Tool WSU-30,
Roll of wire R-30B-0050, (2) 14
DIP’S, (2) 16 DIP’S and Hobby Board
H-PCB-1.
Wire-Wrapping Kit |wK-3B (Blue)| $16.95 1
WIRE DISPENSER
■ With 50 ft. Roll of AWG 30
KYNAR* wire-wrapping wire.
■ Cuts the wire to length.
■ Strips 1" of insulation.
■ Ref illable (For refills, see above)
Blue Wire
WD-30-B
$3.95
Yellow Wire
WD-30-Y
53.95
White Wire
WD-30-W
$3.95
Red Wire
WD-30 R
-$3^r
WIRE-WRAPPING KIT
Contains: Hobby Wrap Tool WSU-30 M ,
Wire Dispenser WD-30-B, (2) 14 DIP’S,
(2) 16 DIP’S, Hobby Board H-PCB-1,
DIP/IC Insertion Tool INS-1416 and
DIP/ 1C Extractor Tool EX-1
Wire-Wrapping Kit |WK-4B (Blue)| $25.99 1
PRE CUT
PRE STRIPPED WIRE
Wire for wire
wrapping.AWG-30
(0.25mm) KYNAR*
wire, 50 wires per
package stripped
1” both ends.
30 AWG blue Wire 1" Lour
30 B 50 010
$99
30 AWG Yellow Wire 1" tonR
30 Y 50 010
$99
30 AWG White Wire. 1 LonR
30 W 50 010
$99
30 AWG Red Wire 1" Long
30 R 50 010
$99
30 AWG Blue Wire ?' Long
30 B 50 020
$107
30 AWG Yellow Wire Lor»R
30 Y 50 020
$1.07
30 AWG White Wire. 2" Long
30 W 50 020
$1.07
30 AWG Red Wire 7 Lotr
30 R 50 020
$1.07
30 AWG Blue Wire 3 Loor
30 8 50 030
$1 16
30 AWG Yellow Wire 3" Loor
30 Y 50 030
$1 16
JU AWG White Wire. 3 LonR
30 W 50 030
$1.16
30 AWG Red Wire. 3" Long
30 R 50 030
$1.16
0
>
%
•jo
%
?
1
30 B 50 040
_ $1 23
30 AWG Yellow Wire 4" Long
30 Y 50 040
$123
30 AWG White Wire 4 Long
30 W 50 040
$123
30 AWG Red Wire 4" Long
30 R 50 040
$1 23
30 AWG Blue Wire 5' Long
30 B 50 050
$1 30
30 AWG Yellow Wire 5" Long
30 Y 50 050
$1 30
30 AWG White Wire. 5" Long
36 W 50 050
$1 30
30 AWG fted Wire. *>" Long
30 R- 50 050
$1 30
30 AWG Blue Wire. 6" Long
30 B 50 060
$1 38
30 AWG Yellow Wire 6" Long
30 Y 50 060
$1 38
30 AWG White Wire 6 ' Long
30 W 50 060
$1 38
30 AWG Red Wire. 6 ' Long
30 R 50 060
$1 38
- KYNAR PtNNWALT
MINIMUM ORDER $25.00, SHIPPING CHARGE SI .00, N Y. CITY ANO STATE RESIDENTS ADD TAX
OK MACHINE & TOOL CORPORATION
3455 Conner St . Bronx. NY 104751(212) 994-6600 ■ Telex 125091
74
DIP/ 1C EXTRACTOR TOOL
The EX-1 Extractor is ideally suited for hobbyist or
lab engineer. Featuring one piece spring steel con
struction. It will extract all LSI. MSI and SSI devices
of from 8 to 24‘pins.
Extractor Tool
EX-1
I * 1 - 49 1
P.C. BOARD
The 4x4.5x1/16 inch board is made of glass coated EPOXY Laminate
and features solder coated 1 oz. copper pads. The board has provision
for a 22/44 two sided edge connector, with contacts on standard .156
spacing. Edge contacts are non-dedicated for maximum flexibility.
The board contains a matrix of .040 in. diameter holes on .100 inch
centers. The component side contains 76 two-hole pads that can accom-
modate any DIP size from 6-40 pins, as well as discrete components.
Typical density is 18 of 14-Pin or 16-Pin DIP'S. Components may be
soldered directly to the board or intermediate sockets may be used for
soldering or wire-wrapping.
Two independent bus systems are provided for voltage and ground on
both sides of the board. In addition, the component side contains 14
individual busses running the full length of the board for complete wir-
ing flexibility. These busses enable access from edge contacts to distant
components. These busses can also serve to augment the voltage or
ground busses, and may be cut to length for particular applications.
Hobby Board
H PCB
i-1 | $4.99]
PC CARD GUIDES
TR-I consists of 2 guides precision molded with
unique spring finger action that dampens shock and
vibration, yet permits smooth insertion or extrac
tion. Guides accommodate any card thickness
from .040 .100 inches.
QUANTITY - ONE PAIR (2 pcs.)
Card Guides
TR-1 $1.89
PC CARD GUIDES & BRACKETS
TRS-2 kit includes 2 TR 1 guides plus 2 mounting
brackets. Support brackets feature unique stabilizing
post that permits secure mounting with only 1 screw.
QUANTITY - ONE SET (4 pcs.)
Guides & Brackets | TRS-2 | $3.79 1
QUANTITY - ONE SET (4 pcs.)
PC EDGE CONNECTOR
44 Pin, dual read out, .156" (3,96
mm) Contact Spacing, .025" (0,63
mm) square wire-wrapping pins.
P.C.. Edge Connector | CON-1 1 $3.49 1
P.C.B. TERMINAL STRIPS
The TS strips provide positive screw activated clamp-
ing action, accommodate wire sizes 14 30 AWG (1, 8-0,
25mm). Pins are solder plated copper. .042 inch (1mm)
diameter, on .200 incn (5mm) centers.
4-Pole
TS- 4
$1.39
8-Pole
TS- 8
$1.89
12-Pole
TS-12
$2.59
DIP SOCKET
Dual-in-line package, 3 level wire-
wrapping, phosphor bronze contact,
gold plated pins .025 (0,63mm) sq.,
.100 (2,54mm) center spacing.
14 Pin Dip Socket
14 Dip
$0.79
16 Pin Dip Socket
16 Dip
$0.89
RIBBON CABLE ASSEMBLY
SINGLE ENDED
With 14 Pin Dip Plug
24" Long (609mm)
SE14-24
$3.55
With 16 Pin Dip Plug
24" Long (609mm)
SE16-24
$3.75
DIP PLUG WITH COVER
FOR USE WITH RIBBON CABLE
14 Pin Plug & Cover
14-PLG
$1.45
16 Pin Plug & Cover
16-PLG
$1.59
QUANTITY: 2 PLUGS. 2
COVERS
RIBBON CABLE ASSEMBLY
DOUBLE ENDED
With 14 Pin Dip Plug -2" Long
DE 14-2
$3.75
With 14 Pin Dip Plug -4" Long
DE 14-4
$3.85
With 14 Pin Dip Plug -8" Long
DE 14-8
$3.95
With 16 Pin Dip Plug -2" Long
DE 16-2
$4.15
With 16 Pin Dip Plug -4" Long
DE 16-4
$4.25
With 16 Pin Dip Plug -8" Long
DE 16-8
$4.35
TERMINALS
■ .025 (0,63mm) Square Post
■ 3 Level Wire-Wrapping
■ Gold Plated
Slotted Terminal
WWT-1
$2.98
Single Sided
Terminal
WWT 2
$2.98
1C Socket Terminal
WWT-3
$3.98
Double Sided
Terminal
WWT-4
$1.98
25 PER PACKAGE
TERMINAL INSERTING TOOL
For inserting WWT-1, WWT-2, WWT-3,
and WWT-4 Terminals into .040
(1,01mm) Dia. Holes.
| INS-1 | $2.49 |
WIRE CUT AND STRIP TOOL
Easy to operate place wires (up to 4) in stripping slot with
ends extending beyond cutter blades . . press tool and pull
. wire is cut and stripped to proper “wire-wrapping" length.
The hardened steel cutting blades and sturdy construction of
the tool insure long life.
Strip length easily adiustable for your applications.
DESCRIPTION
MODEL
NUMBER
ADJUSTABLE
“SHINER’' LENGTH
OF STRIPPED WIRE
INCHES TO INCHES
Price
24 ga. Wire Cut and Strip Tool
ST-100-24
IK." IX."
$ 8.75
26 ga. Wire Cut and Strip Tool
ST-100-26
IK." 1%"
$ 8.75
26 ga. Wire Cut and Strip Tool
ST-100-26-875
7 /b" 1W"
$ 8.75
28 ga. Wire Cut and Strip Tool
ST-100-28
7 /e" lVa"
$11.50
30 ga. Wire Cut and Strip Tool
ST-100-30
7 /a" 1V 8 "
$11.50
THE ABOVE UST OF CUT AND STRIP TOOIS ARE NOT APPUCABIE FOR MYLENE OR TFFI0N INSULATION
OK MACHINE & TOOL CORPORATION
3455 Conner St.. Bronx. N Y 1 0475 ■ (212) 994 -6600* Telex 125091
75
The North Star
Floppy System
an 11 -year-old can build it!
Howie DiBlasi
Director, Vocational Education
Lake Havasu High School
Lake Havasu AZ 86403
m My name is Mark; I am
■ 11 years old. I just
finished a North Star Floppy
Disk Kit. It was easy; I really
made it. And guess what? It
worked the first time I hooked it
up!”
Mark looked at me and
smiled. He was really proud of
himself, and I was too. If an
11-year-old can put the North
Star Kit together, so can you.
“Hey, Dad, am I going to be
rich and famous because I put
the North Star together and you
are writing about me in
Kilobaud ?”
I laughed. Rich? No.
Famous? No. Proud and
satisfied? Yes.
Here We Go
I ordered the North Star Kit
and received it in a week from
the Byte Shop in Phoenix.
When I opened the box and ex-
amined the contents, I was im-
pressed with the quality of the
circuit boards and parts. All the
parts were there, and complete
instructions were included.
After looking over the in-
struction manual, I had my son
read it to see if he understood
what to do. He said, “No
sweat,” and at that point I
decided to let him go ahead and
build the kit.
Printed Matter
Four instruction manuals
came with the kit: (1) Minifloppy
Diskette Storage Drive OEM
Manual; (2) North Star Disk
Operating System Manual; (3)
North Star BASIC Manual; (4)
North Star MICRO-DISK
SYSTEM MDS-A Instruction
Manual.
The instruction manual is
divided into three sections:
theory of operation, assembly
instructions and system in-
tegration and schematics. The
manuals are all well written and
detail numerous situations and
how to set things up. It was a
pleasure to read through and
understand the material. Right
on, North Star!
Assembly
All parts were checked off by
Mark, which helped him
become familiar with the parts
and learn their use. As he
checked them, I took a few
moments to explain the func-
tion of the various parts.
Everything was there. Some
kits don’t always include all
items; but North Star has it all
together.
Mark installed the 47 1C
sockets and soldered them in
place. He had soldered a few
times before so he was familiar
with the correct circuit-board
soldering procedure. He had a
few problems with bridges, but
a little Solder-Wick removed
them. I was pleased to see a
very professional soldermasked
board; properly soldermasking a
board helps to eliminate
problems.
The eight resistors and 40
What you see is what you get. The kit comes complete for the North Star Disk System. The Shugart
disk drive (back right) comes complete and assembled.
76
WOW! Five volts. After the power supply was completed , the con-
nector plug was checked for correct voltages. All OK.
capacitors were then soldered
to the board, and the crystal, 5
volt regulator and heat-sink
hardware followed. It was now
necessary to solder a 34 pin
cable connector to the board.
The MDS Controller board was
plugged into the computer.
Holding his breath, Mark con-
nected the meter, which read +
5 volts. So far so good.
1C Installation
Mark watched while I
demonstrated the correct way
to install the ICs in the sockets.
I made a quick check to make
sure he had them in the correct
location. The manual then gave
two detailed pages of instruc-
tion for waveforms on a scope.
Since I did not have a scope
available we skipped this step.
Power Board Assembly
The disk drive can receive
power three ways: (1) From +5
and + 12 volts from an existing
power supply; (2) power PC
board to regulate power from
an existing unregulated power
supply; (3) North Star power-
supply option (MDS-PS).
Since I knew we would be us-
ing the North Star with two dif-
ferent computers from one time
to another, I had purchased the
North Star power-supply op-
tion. Mark mounted the
transformer in the cabinet and
hooked up the wires, switch
and fuse to complete the power
supply. Ready to test the power
supply for +5 and +12 volts at
the power plug, Mark hooked
up the meter and checked for
the proper voltages. To our
Disk drive assembly. The power supply is assembled to the disk
drive assembly with two spacers and screws on each side. The unit
is then connected to the case.
Look Dad, I did it! A very proud young man. If he can build the
North Star System, you can. Let's go.
satisfaction, they were OK.
The last thing to do was to
make two trace cuts on the
MDS controller board and in-
stall two jumper wires. Done!
Final Check
The real test was drawing
near. Mark installed the MDS
controller board in the com-
puter and hooked up the
cables. With the power switch
and computer on and the disk
in the disk drive, Mark typed EX
E900 and hit return. As he did
that I explained that an asterisk
on the screen signaled that
everything was OK. The next
command was GO BASIC. Mark
did that and BASIC was loaded
in 2 seconds. READY appeared
on the screen and we were
ready to program.
Up And Running
Mark and I input a small pro-
gram to make sure everything
was OK. It was. We sat at the
computer for over three hours
inputting programs and run-
ning them. It was getting late,
so we stopped and decided to
input some more programs dur-
ing the next few days.
Summary
Total construction time for
the project was 4 hours and 20
minutes. You could probably
complete it in less time if you
have experience building kits.
Mark took his time building the
kit, but the time spent paid off
because the system worked the
first time.
While Mark was running a
few programs, I looked over the
manuals. North Star BASIC is
an extended version and has
numerous functions. It also has
an edit function to correct er-
rors; it is a joy to use.
The OEM manual gives com-
plete and detailed description
of the disk drive and complete
schematics. The North Star
Disk Operating Systems
Manual features complete in-
structions and operations for
the DOS. It contains descrip-
tions on creating files, types of
files, deleting files, jump
routines, read and write and
many more procedures that are
available for use. All the
manuals are written so you can
understand them. Maybe some
other manufacturers will take a
lesson from North Star.*
77
A Simple Mailing System
program does all the patching
using POKE statements.
Unfortunately, the advan-
tages of one method of saving
your mailing list over another
are overshadowed by speed
and tape storage problems with
your unit.
a money-making time-saver
Is Speed Your Thing?
Although somewhat slower
than a 250K bit/second floppy-
disk transfer, the lowly cas-
sette is still a good medium for
Stephen Gibson
PO Box 38386
Los Angeles CA 90038
O ne of the first tasks a
small businessman wants
his new computer to do is han-
dle the company mailing list. A
review of the many programs
available reveals a big problem:
Unless you have a disk storage
system, you are forever con-
demned to load all those
names and addresses via the
DATA statement.
To read out the list or print
labels, the data is usually read
into a set of variables, then for-
matted to fit your particular
hard-copy printer. If you fill
your memory or want separate
lists, you have to write a whole
new program. To update, you
are forced to list the program to
find where the last DATA state-
ment ended, then change the
read routine.
All this nonsense takes valu-
able time and makes you a
slave to the machine. It would
be easier to write the program
only once, and simply change
the lists. Here are two ways to
do it: (1) the cassette method
and (2) the bare-bones method.
At least one is bound to work
for you.
Sneaky Software
Secrets Revealed
The problem is not how to
structure the ideal list program
in BASIC, but how to save the
names and addresses in a lan-
guage that doesn’t know how
to save variables. The main pro-
gram should have to be saved
only once.
Surprisingly enough, a num-
ber of rather clever techniques
have been developed to solve
this problem. One method
breaks down the name, a string
variable, and feeds it to tape as
a series of OUT statements.
Another method uses the tape
interface hard-wired in parallel
with the terminal I/O. Still
The entire system here is an Imsai 8080 with 24K of memory,
ADM-3 terminal, Data Duffer, Teletype ... and one efficient
secretary .
another chooses a software
method by patching BASIC’s
terminal I/O over to the tape in-
terface port, and then outputs
the list via PRINT statements
as though it were the terminal
. . . very clever, because the
saving data for later use.
Several cassette interface sys-
tems are available. They differ
widely with respect to speed.
I picked the Tarbell high-
speed interface and coupled it
to the Data Duffer (see Kilo-
baud, March 1978, “Hear It and
Add these lines to the program to let the computer tell you
when to abbreviate.
12 A = 20 :REM WIDTH OF TTY LABEL
1050 IF LEN(NA$(N)) > A THEN GOSUB 5000 : GOTO 1040
1060 IF LEN(CO$(N)) > A THEN GOSUB 5000 : GOTO 1055
1070 IF LEN(AD$(N)) > A THEN GOSUB 5000 : GOTO 1065
1080 IF LEN(CS$(N)) > A THEN GOSUB 5000 : GOTO 1075
1090 IF LEN(ZP$(N)) > A THEN GOSUB 5000 : GOTO 1085
5000 REM
5005 REM LINE LENGTH ERROR
5010 REM
5015 PRINT:PRINT“LINE TOO LONG! ! ”:PRINT:RETURN
Add these lines to run the program with Mits 3.2 12K BASIC
3020 POKE 1776,110 : POKE 1778,32 : POKE 1784,1
3300 POKE 1776,0 : POKE 1778,128 : POKE 1784,1
4015 POKE 1787,110 : POKE 1789,16 : POKE 1794,111
4020 POKE 1778,0 : POKE 1784,255
4090 POKE 1787,0 : POKE 1789,1 : POKE 1794,1
4095 POKE 1778,128 : POKE 1784,1
Add these lines to modify Mits 3.2 8K BASIC to recognize
leading spaces.
13 SP$ = “ ” :REM A SPACE CHARACTER
4022 POKE 528,54 : POKE 529,32 : POKE 530,35
4023 POKE 531,195 : POKE 532,224 : POKE 533,7
4039 POKE 1171,16 : POKE 1172,2
4041 POKE 1171,224 : POKE 1172,7
4046 IF B$ = (E$ + SP$) THEN 4090
4087 POKE 528,0 : POKE 529,0 : POKE 530,0
4088 POKE 531,0 : POKE 532,0 : POKE 533,0
Make these patches to Mits BASIC if you get hung up in the
Tape Input routine and need to return to command level. All
numbers are hexadecimal.
8K 3.2 12K 3.2
Address
Byte
Address
Byte
04D3
80
06F2
80
04D9
01
06F8
01
04DC
00
06FB
00
04DE
01
06FD
01
04E3
01
0702
01
Fig. 1 .
Mits BASIC patches.
See It!”) as a reliable way to use
cassettes without the hassle of
a seemingly endless wait for a
load or the fear that data was
lost because a switch was off
or a knob twisted the wrong
way. The Tarbell manual sug-
gests a variable-saving method
in which the terminal I/O is soft-
ware patched to the cassette
I/O for a transfer. The routines
in the mailing-list program
make these patches to Mits 3.2
8K BASIC (see Program A). The
normal Mits TTY I/O convention
of status port “0” and data port
“1” is used. Patches to Mits 3.2
12K BASIC are also listed in
Fig. 1. If you don’t have Mits
BASIC or a Tarbell, there’s still
hope; you can use the bare-
bones method described later.
Hard-Copy Hassles
Registration is the key ingre-
dient for alignment of the
labels on your printer. A sprock-
eted feed mechanism is almost
a necessity. Of course, you can
simply cut your labels out with
a large paper-cutter, but the
peel-off-type labels are more
convenient and better looking.
You need the sprocket feed to
make them work properly. You
might even consider custom
labels with fancy artwork or the
company logo.
I had quite a time finding off-
the-shelf labels for my old
sprocket-fed Teletype. Almost
everyone sells ready-made
forms for larger printers. There
are a few companies that spe-
cialize in stock or custom la-
bels from camera-ready artwork
(see accompanying “Sources
for More Information”).
If you do start with a Tele-
type, by all means change the
ribbon! Use a carbon ribbon
rather than the stock cloth one
—the printing looks so much
better. Unique type fonts are
also available for the Teletype.
Even the Teletype can be made
to look as good as an IBM Se-
lectric ... as long as you don’t
mind all caps— not an earth-
shaking problem for a simple
mailing system such as this.
You will have to change the
platen if your Teletype is a fric-
tion-feed model. The modifica-
tion to your machine is simple
and inexpensive. I’m not advo-
cating the Teletype as the ideal
printer for this system; my com-
pany just happens to have one.
Besides being slow, it’s noisy!
Eventually, I had to stick ours
off in a room by itself to drown
out the clatter. The advantages,
of course, are that the machine
is reliable and inexpensive.
Used machines abound, and
service is readily available.
Simple Program Does It All
Only four routines make up
the cassette program. In the
listing in Program A, lines 1 to
50 initialize the program. A gen-
erous 10,000 bytes are cleared
away based on an average line
length of 20 characters, with 5
lines given to each company
and a list size of 100 com-
panies. The variables S and L
represent the maximum size of
the list and the current list size,
respectively. The subscripted
variables in line 25 are dimen-
sioned to the size of the list. Of
course, you can set this value
higher for a larger list if your
memory capacity will permit it.
The command level routine
prints suitable prompts for
those unfamiliar with the pro-
gram. A branch is made at line
155 based on the value of C.
To enter names at line 1000,
the list counter L is in-
cremented by 1 and a test is
made to see if the list size is
greater than 100 names. It
might be later on, so we must
check it out. If so, the list
counter is decremented back to
100 and a return is made to the
command routine. In line 1030,
a message indicates that the
number symbol (#) can be used
to exit the routine. A FOR /
NEXT loop inputs the names
and addresses into the sub-
scripted variables.
You might wish to make the
prompts different for your ver-
sion. Instead of “ZIP . . . ,” for
instance, you might want the
program to print “COUN-
TRY . . . ,” if you mail overseas.
Or you could eliminate “ZIP”
(ZP$) altogether and squeeze it
into the CITY/STATE line.
If # is typed in line 1045, a
branch is made and the list
1 REM
2 REM **** MAILING LIST ****
3 REM
4 REM BY STEPHEN GIBSON 1/10/77
5 REM RUNS ONLY ON MITS 3.2 8K BASIC
6 REM AND TARBELL CASSETTE INTERFACE
7 REM
8 REM INITIALIZE
9 REM
10 CLEAR 10000 :REM CLEAR SPACE FOR LIST
15 S = 100 :REM MAXIMUM LIST SIZE
20 L = 0 :REM CURRENT LIST SIZE
25 DIM NA$(S),CO$(S),AD$(S),CS$(S),ZP$(S)
30 E$ = :REM END OF LIST CHARACTER
35 OUT 1,26 :REM CLEARS SCREEN
40 PRINTTAB(20);“*** THIS IS MAILING LIST ***”
50 PRINT
100 REM
105 REM COMMAND LEVEL ROUTINE
110 REM -
115 PRINT“PLEASE ENTER YOUR COMMAND: ”:PRINT
120 PRINT“ENTER NAMES INTO LIST = 1”
125 PRINT“PRINT-OUT OF LIST =2”
130 PRINT“STORE LIST ON TAPE = 3’’
135 PRINT“READ LIST FROM TAPE = 4”
140 PRINT
145 INPUT“COMMAND”;C
150 IF 04 THEN 115
155 ON INT(C) GOTO 1000, 2000, 3000, 4000
1000 REM
1005 REM ENTER NAMES ROUTINE
1010 REM
1015 L = L+ 1
1020 IF L > 100 THEN 1400
1025 PRINT“IF YOU WISH TO EXIT THIS ROUTINE ...”
1030 PRINT“TYPE ONE OF THESE ‘#’,THEN ‘RETURN’.”
1035 FOR N = L TO 100 :PRINT:PRINT“NUMBER ”;N:PRINT
1040 INPUT“NAME : ”;NA$(N)
1045 IF NA$(N) = “#” THEN 1300
1047 IF NA$(N) = “\” THEN N = N-2 :GOTO 1100
1050 REM
1055 INPUT“COMPANY : ”;CO$(N)
1060 REM
1065 INPUT” ADDRESS : ”;AD$(N)
1070 REM
1075 INPUT‘‘CITY & STATE : ”;CS$(N)
1080 REM
1085 INPUT‘‘ZIP : ”;ZP$(N)
1100 NEXT
1200 L = 100 : GOTO 1500
1300 L = N - 1 : GOTO 1600
1400 L = L - 1
1500 PRINT:PRINT“THE LIST IS FULL.”:PRINT
1600 PRINT:PRINT“YOU HAVE ”;L;‘‘NAMES ON THIS LIST.”
1700 GOTO 100
2000 REM
2005 REM PRINT-OUT ROUTINE
2010 REM
2015 REM PRINT
2020 PRINT‘‘l) LINE UP LABELS IN PRINTER.”
2025 PRINT
2030 PRINT”2) TURN ON PRINTER.”
2035 PRINT
2040 PRINT* ‘3) TYPE ANY LETTER, THEN ‘RETURN’.”
2045 PRINT:INPUT‘ ‘WAITING . . . ”;W$
2050 FOR X = 1 TO L STEP 3
2055 Y = X+ l:Z = X + 2
2060 PRINT TAB(0) ; NA$(X) ; TAB(25) ; NA$(Y) ; TAB(51) ; NA$(Z)
2065 PRINT TAB(0) ; CO$(X) ; TAB(25) ; CO$(Y) ; TAB(51) ; CO$(Z)
2070 PRINT TAB(0) ; AD$(X) ; TAB(25) ; AD$(Y) ; TAB(51) ; AD$(Z)
2075 PRINT TAB(0) ; CS$(X) ; TAB(25) ; CS$(Y) ; TAB(51) ; CS$(Z)
2080 PRINT TAB(0) ; ZP$(X) ; TAB(25) ; ZP$(Y) ; TAB(51) ; ZP$(Z)
2085 PRINT:PRINT
2090 NEXT
2095 GOTO 100
3000 REM i-
3005 REM STORE ON TAPE ROUTINE
3010 REM
3011 PR1NT:PRINT“1) PLACE NEW CASSETTE IN RECORDER.”
3012 PRINT:PRINT‘‘2) PUT IN RECORD MODE AND ZERO COUNTER.”
3013 PRINT:PRINT‘‘3) WAIT A FEW SECONDS TO ALLOW A LEADER.”
3014 PRINT:INPUT‘‘4) TYPE ANY LETTER, THEN ‘RETURN’. ”;W$
3015 S$ = CHR$(195) + CHR$(230)
3020 POKE 1233,110 : POKE 1235,32 : POKE 1241,111
3025 FOR N = 1 TO L
3030 D$(l) = NA$(N)
3035 D$(2) = CO$(N)
3040 D$(3) = AD$(N)
3045 D$(4) = CS$(N)
3050 D$(5) = ZP$(N)
3055 FOR J = 1 TO 5
3060 FOR K = 1 TO 100 : NEXT K
3065 B$ = S$ + D$(J)
3070 PRINT B$
3075 NEXT J
3080 NEXT N
3085 FOR T = 1 TO 3
3090 B$ = S$ + E$
3095 FOR K = 1 TO 100 : NEXT K
3100 PRINT B$
3200 NEXT T
3300 POKE 1233,0 : POKE 1235,128 : POKE 1241,1
3400 GOTO 100
4000 REM
4005 REM READ FROM TAPE ROUTINE
4010 REM
4011 PRINT: PRINT” 1) PLACE CASSETTE IN RECORDER.”
4012 PRINT:PRINT”2) SET COUNTER AND PUSH PLAY.”
4013 PRINT: PRINT* ‘3) ALLOW TIME FOR LEADER.”
4014 PRINT:INPUT‘‘4) TYPE ANY LETTER, THEN ‘RETURN’.”;W$
4015 POKE 1244,110 : POKE 1246,16 : POKE 1251,111
4020 POKE 1235,0 : POKE 1241,255
4025 FOR N = 1 TO 101
4030 FOR J = 1 TO 5
4035 OUT 110,16
4040 INPUT BS
4045 IF B$ = E$ THEN 4090
4050 D$(J) = B$
4055 NEXT J
4060 NA$(N) « D$(l)
4065 COS(N) = D$(2)
4070 AD$(N) = D$(3)
4075 CS$(N) = D$(4)
4080 ZP$(N) = D$(5)
4085 NEXT N
4090 POKE 1244,0 : POKE 1246,1 : POKE 1251,1
4095 POKE 1235,128 : POKE 1241,1
4100 L = N - 1
4200 PRINT:PRINT”THIS LIST HAS ”;L;“ NAMES ON IT.”:PRINT
4300 GOTO 100
Program A. Program listing for A Simple Mailing System. Here are the routines you need to patch
Mits 8K 3.2 BASIC to load or save your mailing list using the Tarbell high-speed cassette interface.
counter L is decremented by
one (1) and a return is made to
the command routine. Some-
times I make mistakes when en-
tering a name (my secretary
never does). I find it convenient
to be able to type a character
that tells the program to go
back one name and start over.
Line 1047 does it all. I chose a
backslash, but you should feel
free to choose your own char-
acter to personalize this pro-
gram. You could insert this line
after every input if you’d rather
check your work a line at a time.
Another useful addition is to
print a space, for example,
where the name goes in the
event you have a company
name, but no one individual to
mail to. A space is a logical en-
try. Don’t try it unless you add
the appropriate lines from Fig.
1 because Mits BASIC ignores
leading spaces. The listed
POKES change the input rou-
tine to add a space if a carriage
return is received. I found it
convenient to print the current
list size in line 1600 before ex-
iting this routine.
The printout routine must be
tailored to your particular print-
er. The program format given is
for a standard Teletype using
peel-off labels spaced three
across. Lines 2020 to 2045 give
instructions. The variable W$ is
only a buffer to wait until you
are ready to print. Extra PRINT
statements in line 2085 ad-
vance the form to the next set
of labels. To print your labels
three at a time for the popular
machine-gun mailings, simply
substitute the lines in Fig. 2.
The store (on tape) routine at
line 3000 begins the really use-
ful aspects of this program. It is
here that the names and ad-
dresses only are fed to tape. In-
structions are given in lines
3011 to 3014. W$ is still only a
wait buffer. S$ is set to the
value of the Tarbell start and
sync bytes. POKEs to Mits
BASIC are then made in line
3020 to shift the terminal I/O to
the cassette I/O port. The
names and addresses are
placed in a D$ buffer, then out-
put with the start and sync
bytes as B$ via PRINT
statements.
80
Instead of this format . . .
John Craig
Wayne Green
Stephen Gibson
Editor
Publisher
Famous author
Kilobaud Magazine
Kilobaud Magazine
PO Box 38386
Peterborough NH
Peterborough NH
Los Angeles CA
03458
03458
90038
You might want this . . .
John Craig
John Craig
John Craig
Editor
Editor
Editor
Kilobaud Magazine
Kilobaud Magazine
Kilobaud Magazine
Peterborough NH
Peterborough NH
Peterborough NH
03458
03458
03458
Wayne Green
Wayne Green
Wayne Green
Publisher
Publisher
Publisher
Kilobaud Magazine
Kilobaud Magazine
Kilobaud Magazine
Peterborough NH
Peterborough NH
Peterborough NH
03458
03458
03458
Then substitute these lines
2050 REM 3-UP FORMAT
2055 FOR N = 1 TO L
2060 PRINT TAB(0);NA$(N);TAB(25);NA$(N);TAB(51);NA$(N)
2065 PRINT TAB(0);CO$(N);TAB(25);CO$(N);TAB(51);CO$(N)
2070 PRINT TAB(0);AD$(N);TAB(25);AD$(N);TAB(51);AD$(N)
2075 PRINT TAB(0); CS$(N);TAB(25); CS$(N);TAB(51); CS$(N)
2080 PRINT TAB(0); ZP$(N);TAB(25); ZP$(N);TAB(51); ZP$(N)
Fig. 2. Instead of this format
The delay loop in line 3060
bears some explanation. When
data is brought back into the
program, allow time for BASIC
to reinsert the data into the ap-
propriate subscripted variables
by implementing a delay during
the output sequence. You
could, perhaps, shorten the
delay, but you might lose some
of your data. A value of lOOforT
allows plenty of safety.
The End of List character, E$,
must also be output. The com-
puter will look for this charac-
ter when the list is played back
into the machine to set the list
counter. This particular ar-
rangement allows lists of vary-
ing size and the addition of
more names to a short list.
Beginning on line 3090, E$ is
linked to the start and sync
bytes and output three times.
Why three; isn’t once enough?
That’s right. But suppose you
had a dropout on the tape. It
does happen on old cassettes,
particularly cheap ones. Even if
you use top-notch cassettes,
you may still lose a byte be-
cause your recorder’s slow
AGC attack time may turn the
beginning of a byte to garbage.
I proved it writing this program.
The computer missed the E$
on playback. It just sat there
waiting. It was very annoying
. . . especially because the pro-
gram had POKEd the I/O away
from my terminal to the cas-
sette interface. I had no way to
talk to my machine except via
the system monitor and the
front panel to patch things up
between my computer and its
program. The pandemic Mur-
phy’s Law says you won’t need
to use the patches I made if I
list them in Fig. 1. I output the
E$ three times, rather than
once, and beat old Edsel Mur-
phy by even a New York sec-
ond! (That’s easy for me to say,
you say.) The routine ends by
POKEing BASIC back to normal
I/O and jumping to the com-
mand routine.
The tape input routine is very
similar. Instructions are given
in lines 4011 to 4014. The I/O is
POKEd to the cassette port just
as before, and data is input by
another FOR/NEXT loop. It is
useful to print out the size of
the list after the I/O is POKEd
back because not all lists will
be set at the maximum size.
You will then be able to add to
the current list by using the in-
put routine. Then save the
whole thing as a full list.
The Bare-Bones Method
Suppose you have a compu-
ter and a Teletype, but neither
speaks Mits BASIC nor recog-
nizes Tarbell format. If your
Teletype has a paper-tape
punch (most do), you can still
benefit from this system.
Start by making those nifty
mods to the Teletype, especial-
ly the ribbon. Then enter the
program in Program B. The vari-
ables are the same as the cas-
sette program, but the prompts
are different and the save and
read routines are left out.
Next, run the program and
enter the names and ad-
dresses. When you print the
list, simply turn on the paper-
tape punch at the same time.
You will have an exact copy of
the printout, as well as a set of
labels, on paper tape. You can
then reprint the list by using the
Teletype in the local mode and
reading off the paper tape. Turn
on the punch again while print-
ing if you need a spare copy of
your list. Use a separate punch
if you have one.
2 REM **** BARE BONES MAILING LIST ****
3 REM
4 REM BY STEPHEN GIBSON 12/11/76
5 REM RUNS ON ASR-33 TTY OR SIMILAR
6 REM PRINTER WITH PAPER TAPE PUNCH
7 REM
8 REM INITIALIZE
9 REM
10 CLEAR 10000 :REM CLEAR SPACE FOR LIST
15 S= 100 :REM MAXIMUM LIST SIZE
20 L = 0 :REM CURRENT LIST SIZE
25 DIM NA$(S),CO$(S),AD$(S),CS$(S),ZP$(S)
30 E$ = :REM END OF LIST CHARACTER
35 OUT 1,26 :REM CLEARS SCREEN
40 PRINTTAB(20);“*** THIS IS MAILING LIST ***”
50 PRINT
100 REM
105 REM COMMAND LEVEL ROUTINE
110 REM
115 PRINT“PLEASE ENTER YOUR COMMAND : ’ * : PRINT
120 PRINT“ENTER NAMES INTO LIST = 1”
125 PRINT“PRINT-OUT OF LIST = 2”
140 PRINT
81
145 INPUT“COMMAND”;C
150 IF 02 THEN 115
155 ON INT(C) GOTO 1000, 2000
1000 REM
1005 REM ENTER NAMES ROUTINE
1010 REM
1015 L = L+ 1
1020 IF L > 100 THEN 1400
1025 PRINT“IF YOU WISH TO EXIT THIS ROUTINE ...”
1030 PRINT‘‘TYPE ONE OF THESE ‘#’,THEN ‘RETURN’.”
1035 FOR N = L TO 100 :PRINT:PRINT‘‘NUMBER ”;N:PRINT
1040 INPUT‘‘NAME : ”;NA$(N)
1045 IF NAS(N) = “#” THEN 1300
1047 IF NA$(N) = “Y* THEN N = N-2 :GOTO 1100
1050 REM
1055 INPUT 4 ‘COMPANY : ”;CO$(N)
1060 REM
1065 INPUT 44 ADDRESS : ”;AD$(N)
1070 REM
1075 INPUT‘‘CITY & STATE : ”;CSS(N)
1080 REM
1085 INPUT“ZIP : ”;ZP$(N)
1100 NEXT
1200 L = 100 : GOTO 1500
1300 L = N - 1 : GOTO 1600
1400 L = L - 1
1500 PRINT:PRINT“THE LIST IS FULL.”:PRINT
1600 PRINT:PRINT“YOU HAVE ”;L; 44 NAMES ON THIS LIST.”
1700 GOTO 100
2000 REM
2005 REM PRINT-OUT ROUTINE
2010 REM
2015 PRINT
2020 PRINT‘‘l) MAKE PAPER TAPE LEADER IN ‘LOCAL’ MODE.”
2025 PRINT
2030 PRINT“2) SWITCH PRINTER TO ‘LINE’ AND LINE UP LABELS.”
2035 PRINT
2040 PRINT‘‘3) TYPE ANY LETTER, THEN ‘RETURN’.”
2045 PRINT:INPUT“ WAITING . . . ”;W$
2050 FOR X = 1 TO L STEP 3
2055 Y = X + 1 : Z =X + 2
2060 PRINT TAB(0) ; NA$(X) ; TAB(25) ; NA$(Y) ; TAB(51) ; NA$(Z)
2065 PRINT TAB(0) ; CO$(X) ; TAB(25) ; COS(Y) ; TAB(51) ; CO$(Z)
2070 PRINT TAB(0) ; AD$(X) ; TAB(25) ; ADS(Y) ; TAB(51) ; AD$(Z)
2075 PRINT TAB(0) ; CS$(X) ; TAB(25) ; CS$(Y) ; TAB(51) ; CSS(Z)
2080 PRINT TAB(0) ; ZP$(X) ; TAB(25) ; ZP$(Y) ; TAB(51) ; ZP$(Z)
2085 PRINTrPRINT
2090 NEXT
2095 GOTO 100
Program B. Listing for the bare-bones version of the program. The format is set for a Teletype. Sim -
pie adjustments can be made to fit other printers. A paper-tape punch is used to save the list. The
Teletype is run in local mode to print additional lists.
We don’t have to confine our
list to names and addresses.
Adding a few more variables in
the program allows the luxury
of obtaining other important
data from our list, such as
types of merchandise each cus-
tomer wants or has ordered.
You might choose to save im-
portant dates for each custom-
er— write a simple routine to
search the current list and pop
out names that need collection
letters, birthday greetings or
warranty follow-up letters. The
personalized form letter, men-
tioned before, could be printed
just for those on the list who
need it. All you need do is add
to the routines given.
Perhaps you can begin to see
that what started as a simple
system could easily be expand-
ed into a first-class data base
for your business. You can start
with the program given and up-
grade from there, even to disk.
You lose nothing by starting
now with just the list. In fact,
you may gain in the long run be-
cause you will be able to tailor
the program to your own needs.
The really important proce-
dures will be yours, thereby
ending forever that locked-in
feeling you get with someone
else’s software.
If you know that feeling or
need an upward compatible
mailing program for your busi-
ness, you should get this pro-
gram up and running and begin
to save time and money now
while planning for the future. ■
This particular method is in-
expensive and does not take
any time at all to load or make
because the paper-tape copy is
punched as you print the list!
How easy can something be?
If It Works . . . Modify It!
Suppose your names are
longer than your labels. When
do you abbreviate? Adding the
appropriate lines from Fig. 1
allows the computer to count
the number of input characters.
The LEN function, if you have it
(Mits does), can test against
the size of your line. If the test
is valid, GOSUB to an error mes-
sage. Further modifications in-
clude another module to read a
whole letter from cassette us-
ing the POKES given. Then print
a personalized copy to each
customer on the list.
You can now consider sen-
tence structures like “and in
closing, ‘Mr. Jones,’ we’d like to
offer . . . just as the big mail-
order operations do it! Still an-
other useful modification is a
cassette tape directory of your
lists ... a good idea when you
get up to a thousand. An ex-
cellent example of this method
appeared in 73 Magazine (“The
Soft Art of Programming,”
Parts 1-3, Oct-Dec 1976, by Rich
Didday) and was reprinted in
The New Hobby Computers, 73
Inc., 1977.
Sources For More Information
High-speed cassette
Tarbell Electronics
interface.
20620 S. Leapwood Ave. Suite P
Carson CA 90746
(213) 538-4251
Teletype labels and
Uarco Incorporated
ready-made forms for
2600 Wilshire Bl. Suite 408
printers.
Los Angeles CA 90057
(213) 380-2595
Custom labels for any
Avery Label Company
printer.
777 E Foothill Bl.
Azusa CA 91704
(213) 969-3311
Teletype sprocket feed
TTS
kits and special type
2928 Nebraska Ave.
fonts. Also carbon
Santa Monica CA 90404
ribbons for Teletype.
(213) 829-2611
82
THE SSB *150
FLOPPY DISCOUNT
Affordable
The tribe at Smoke Signal Broadcasting took our
BFD-68 disk system and scalped the price, but
not the features to create the ABFD-68 (Affordable
Basic Floppy Disk). We appreciate the fact that
the computer hobbyist gave us our start and we
haven’t forgotten you.
$649 Assembled
Compare Price. Our SS-50 bus compatible disk
system is $1 50 less than the assembled price of the
leading S-1 00 disk system. And you can at least
double that savings when you buy one of the
computers manufactured by MSI or SWTPC that
use the superior 6800 microprocessor.
Programmable
The BFD-68 is well known for its fine software. The
system comes with the best disk operating system
available and we offer a multitude of other com-
patible software products. These include a BASIC
interpreter with disk file handling capability. By
the way, our DOS now easily handles true random
access files as well as sequential. Also, we have a
super fast BASIC compiler for business applica-
tions. In addition, a Text Editor, 2 Assemblers, a
Trace Disassembler useful for program debugging
and an Object to Source Code Generator are all
stock items available for immediate delivery.
A word processor will be available very soon.
Reliable
We delivered our first mini-floppy disk system a
year ago — 6 months ahead of any other 6800
based mini system. Thus, we’ve had twice the
experience in building reliability into the system.
Our NEW disk controller was designed using all
we have learned in the past year about system
reliability.
The ABFD-68 contains all the built in reliability
of our regular BFD-68 plus you save money by
supplying your own cabinet and power supply
for the disk.
Available
We’ve shipped literally tons of our BFD-68 disk
system in the past year and have learned to keep
our production up with demand. Give us a call and
chances are we’ll be able to ship you the new
ABFD-68 from stock and charge it to your Master
Charge or Visa card. Better yet, ask us for the name
of the computer store nearest you that carries our
complete line of computer products.
SMOTE SUM BBMMffllK
P.O. Box 20 1 7, Hollywood, CA 90028 *(213) 462-5652
S46
83
Dr. Adam Osborne
Osborne & Associates, Inc.
PO Box 2036
Berkeley CA 94702
Number Crunching:
Two Hardware Solutions
faster and smoother than software
P eople attempting to use
microprocessors in scienti-
fic applications are probably
the first to discover that micro-
processors do indeed have limi-
tations. A microprocessor’s
ability to execute instructions
in microseconds may, on the
surface, sound very impressive,
and it is— until you try to han-
dle trigonometric functions,
logarithms, exponentiation or
even multidigit multiplication
and division.
Trigonometric and logarith-
mic functions are generally re-
ferred to as “transcendental”
functions. Writing a microcom-
puter program to handle trans-
cendental functions is far more
difficult than the most complex
payroll system could ever be. In
fact, designing a program that
will generate truly accurate
transcendental functions is a
formidable task. The problem
with these functions is that
over limited ranges they
change rapidly. Programs that
generate transcendental func-
tions must generate very ac-
curate answers, particularly in
the fast-moving range, because
on rare occasions you will want
to subtract almost identical val-
ues— and a small difference be-
tween two large, erroneous
numbers may be completely
wrong.
Therefore, when examining
arithmetic processors, you
must look at the accuracy of
results in the fast-moving
numeric range. If you are ac-
customed to evaluating chips
simply on the basis of cycle
times and programmable op-
tions, you now have an impor-
tant new consideration— the
method used to generate
results.
Two arithmetic processors
will soon be available: the
MM57109 from National Semi-
conductor and the AM9511
from Advanced Micro Devices.
About the only thing these two
devices have in common is that
they both perform approx-
imately the same transcenden-
tal functions, and each is
treated as a support device
within a microcomputer
system.
Suppose, for example, you
want to compute the natural
logarithm of a number. You will
transmit the number, as data,
to an arithmetic processor, ad-
dressing it as an I/O port. At
some later time you will read
back the answer, as data being
input from an I/O port. This use
of an arithmetic processor is il-
lustrated conceptually in Fig. 1.
The primary difference be-
tween the MM57109 and the
AM9511 is that National Semi-
conductor’s MM57109 is a cal-
culator chip; it looks nothing
II ADDRESS BUS
I { DATA BUS
I{ CONTROL SIGNALS
Fig. 1. An arithmetic processor in a microcomputer system.
INSTRUCTION
CODES (6 BIT)
0
c
CONTROL
LOGIC
CONTROL SIGNALS
8 BCD DIGIT
REGISTERS
k DIGIT ADDRESS
(4 BIT)
, DIGIT OUTPUT
(4 BIT)
Fig. 2. MM57109 arithmetic processor functional logic.
84
INT
READY
ACKNOWLEDGE
PREVIOUS
OPERATION
COMPLETED
BY MM57I09
r— NEW INSTRUCTION ACCEPTED
j BY MM57I09
♦
*] (OUTPUT FROM
MM57I09)
| (INPUT TO
| MM57I09)
MICROPROCESSOR TRANSMITS
NEW INSTRUCTION TO MM57I09
MM57I09 WAITS
FOR NEXT
INSTRUCTION
READY
Fig. 3.
Fig. 4.
like the typical microprocessor
support device. The AM9511, in
contrast, is immediately
recognizable to any ex-
perienced microcomputer user
as a typical microprocessor
support device.
Let’s look at each part in
turn. The discussion that
follows will give you some idea
of part capabilities; however,
detailed operating procedures
are not provided.
MM57109
Fig. 2 illustrates the general
logic organization of the
MM57109. The most important
characteristic of this part is
that it operates on binary-
coded decimal (BCD) numbers
up to eight digits long.
Numbers may be handled in
fixed-point or floating-point for-
mat. A fixed-point number is
eight digits long, with adecimal
point located at any digit boun-
dary. Thus, numbers in the
range 99999999 through
.00000001 may be represented.
Floating-point numbers have
the form:
( ± O.XXXXXXXX)ExP( ± YY)
X and Y represent any decimal
digits. Thus, any number in the
range 1 x 10 + 99 through 1 x
10 ”99 can t> e represented,
with eight digits of accuracy.
As you might expect, you
must operate the MM57109 by
transmitting data and com-
mands to it. Results are re-
ceived as data. Commands are
summarized in Table 1. Note
that the ^/IM57109 is not a fast
device.^ Execution times are
shown based on a ten-micro-
second microcycle, the recom-
mended maximum rate for this
device. It takes at least four
milliseconds to enter a single
eight-digit number (in fixed- or
floating-point notation), while
trigonometric functions may
take almost a second to
resolve.
In order to cope with these
relatively slow times, all data
communications between the
MM57109 and a microproces-
sor use request/acknowledge
handshaking control signal
protocol. Upon completing any
operation, the MM57109 out-
puts a ready signal true. Nor-
mally the microprocessor will
hold an acknowledge input
false to suppress any new
operations. Upon detecting the
true ready, the microprocessor
will transmit a new command
to the MM57109 and set the
acknowledge input true. This is
illustrated in Fig. 3.
This handshaking scheme
readily lends itself to almost
any microprocessor; the ready
“true” signal can be used to re-
quest an interrupt, while the
acknowledge can be tied direct-
ly to a combined MM57109 de-
vice-select and write-control
signal. For 8080A signals, this
is illustrated in Fig. 4.
The method of transmitting
control commands to an
MM 571 09 device differs mark-
edly from the standard method
used within microcomputer
systems. The standard method
(which is used by the AM9511)
takes the device-select logic
output to a select pin, then has
a control/data discriminator
that usually constitutes part of
the device address. Memory-
read and memory-write control
signals then become simple
control strobes that accom-
pany an address-activated
select logic. Fig. 5 illustrates
this.
There are three ways you can
enter data to the MM57109; in
each case the register stack is
pushed and data is written into
the X register (see Fig. 6).
The first data entry method is
approximately equivalent to
calculator-keyboard entry; sep-
arate commands identify the
Table 1. MM57109 instruction description table (* indicates two-word instruction).
EXECUTION
EXECUTION
TIME
TIME
(MICROCYCLES)
(MICROCYCLES)
(WORST-CASE
MNEMONIC* (AVERAGE)
VALUES)
DESCRIPTION
0
238
Mantissa or exponent digits. On first digit (d) the
1
238
following occurs: Z— T
2
238
Y— Z
3
238
X-Y
4
238
d— X
5
238
See description of number entry on page 11.
6
238
7
238
8
238
9
238
DP
152
Digits that follow will be mantissa fraction.
EE
151
Digits that follow will be exponent.
CS
166
Change sign of exponent or mantissa.
Xm = X mantissa
Xe = X exponent
CS causes -Xm— Xm or -Xe— Xe depending on
whether or not an EE instruction was executed
PI
1312
after last number entry initiation.
EN
552
3.1415927— X, stack not pushed.
Terminates digit entry and pushes the stack.
The argument entered will be in X and Y.
Z-T
Y-Z
X-Y
NOP
122
Do nothing instruction that will terminate digit
entry.
HALT
134
External hardware detects HALT op code and
generates HOLD = 1. Processor waits for HOLD
= 0 before continuing. HALT acts as a NOP and
may be inserted between digit entry instructions
since it does not terminate digit entry.
ROLL
905
Roll Stack.
T Y
85
POP
448
XEY
652
XEM
812
MS
839
MR
1385
LSH
168
RSH
173
+
2200
6600
2200
6600
X
3200
22700
/
7800
22300
YX
55400
95500
INV + *
1700
5000
INV - *
1700
5000
INV x *
2700
21400
INV/*
7300
21100
1/X
4500
22800
SORT
7000
30200
SQ
3000
21900
10X
27400
96500
EX
30800
93900
LN
24800
92000
LOG
30700
92600
SIN
56200
95900
COS
56200
95900
TAN
35000
97600
INV SIN*
54000
93900
INV COS*
54000
93900
INV TAN*
30200
92900
DTR
9600
41700
RTD
9600
41700
MCLR
734
ECLR
163
JMP*
186
TJC*
208
TERR*
191
o
II
X
H
278
TXF*
277
TXLT0*
197
IBNZ
2314
DBNZ
2314
IN*
395
Pop Stack.
Y-X
Z-Y
T-Z
O-T
Exchange X and Y.
X«— -Y
Exchange X with memory.
X— *M
Store X in Memory.
X-M
Recail Memory into X.
M— 'X
X mantissa is left shifted while leaving decimal
point in same position. Former most significant
digit is saved in link digit. Least significant digit is
zero.
X mantissa is right shifted while leaving decimal
point in same position. Link digit, which is normal-
ly zero except after a left shift, is shifted into the
most significant digit. Least significant digit is
lost.
Add X to Y. X + Y - X. On +,-,x,/ and YX in-
structions, stack is popped as follows:
Z—Y
T-*Z
O— T
Former X, Y are lost.
Subtract X from Y. Y - X -* X
Multiply X times Y. Y x X -*• X
Divide X into Y. Y ■+• X -* X
Raise Y to X power. Y x — X
Add X to memory. M + X -* M
On INV +, x and / instructions, X, Y, Z, and T
are unchanged.
Subtract X from memory. M - X — M
Multiply X times memory. M x X — M
Divide X into memory. M + X — M
1 + X -*• X. On all F (X) math instructions Y,Z,T and
M are unchanged and previous X is lost.
n/TC- X
X2-X
10 X -X
e x -X
InX-X
log X - X
SIN(X) - X. On all F(X) trig functions, Y,Z,T, and M
are unchanged and the previous X is lost.
COS(X) - X
TAN(X) - X
SIN “ ^(X) -* X
cos-^xj-x
TAN “ 1 (X) - X
Convert X from degrees to radians.
Convert X from radians to degrees.
Clear all internal registers and memory; initialize
I/O control signals, MDC = 8, MODE = floating
point. (See initialization.)
O -* Error flag
Unconditional branch to address specified by sec-
ond instruction word. On all branch instructions,
second word contains branch address to be load-
ed into external PC.
Branch to address specified by second instruc-
tion word if JC (lg) is true ( = 1). Otherwise, skip
over second word.
Branch to address specified by second instruc-
tion word if error flag is true ( = 1). Otherwise, skip
over second word. May be used for detecting
specific errors as opposed to using the automatic
error recovery scheme dealt with in the section on
Error Control.
Branch to address specified by second instruc-
tion word if X = 0. Otherwise, skip over second
word.
Branch to address specified by second instruc-
tion word if |X| < 1. Otherwise, skip over second
word, (i.e., branch if X is a fraction.)
Branch to address specified by second instruc-
tion word If X < 0. Otherwise, skip over second
word.
M + 1 -*■ M. If M = 0, skip second instruction
word. Otherwise, branch to address specified by
second instruction word.
M - 1 -* M. If M = 0, skip second instruction
word. Otherwise, branch to address specified by
second instruction word.
The processor supplies a 4-bit digit address ( DA4 -
DA1) accompanied by a digit address strobe (DAS)
decimal digits 0 through 9, the
decimal point and signs for the
mantissa and exponent— if
floating-point format is
specified.
The other two input tech-
niques transmit data to the X
register under program control.
An IN instruction is executed
once for entry of an entire num-
ber, while an AIN instruction is
executed once per digit of a
number being entered. In each
case the number is entered into
the X register after the stack is
pushed, as illustrated for key-
board entry. Following execu-
tion for the IN or AIN instruc-
tion, digits are entered as data.
Input is clocked by an output
control signal accompanying
the 4-bit digit address il-
lustrated in Fig. 2.
Handshaking protocol simi-
lar to the ready-acknowledge
sequence illustrated for in-
struction input controls data
entry. Thus, it is relatively easy
for any microprocessor to work
asynchronously with the
MM57109.
MM57109 data output is con-
trolled by an OUT instruction
which is equivalent to the IN in-
struction.
MM57109 data input and out-
put philosophy contrast sharp-
ly with normal microprocessor
protocol. Observe that the
MM57109 requires the micro-
processor to input an ap-
propriate control command,
after which the MM57109 out-
puts strobe signals to time data
input or output. Thus, the
MM57109 is not behaving like a
standard peripheral device,
rather, it becomes temporary
bus master while inputting or
outputting data.
In a normal microcomputer
system, the microprocessor
will input or output data from a
support device just as it would
for read/write memory. The
device is selected via an ap-
propriate I/O port or memory
address, then a read or write
control signal causes the data
transfer to occur; this is how
the AM9511 works.
National Semiconductor lit-
erature describes the MM57109
as either a stand-alone micro-
processor or as an adjunct to
another microprocessor. In re-
ality, the MM57109 is not a
practical stand-alone micropro-
cessor. It should be used only
in conjunction with another
microprocessor because the
MM57109 has no internal mem-
ory-addressing logic. A pro-
gram counter, if present, must
be implemented externally, us-
ing some appropriate register
whose contents get triggered
when appropriate timing sig-
nals are output by the
MM57109. Branch instructions,
though identified in Table 1,
really do not exist; they simply
create a control signal that ex-
ternal logic must use to clock
an address into the external
program counter.
By the time you have config-
ured the necessary additional
logic to surround a stand-alone
MM57109, you will probably
find it is cheaper and a good
deal faster to use some simple
microprocessor, even if its sole
function is to monitor and con-
trol MM57109 operations.
AM9511
Now let’s look at the AM951 1 .
Functional logic for this device
is illustrated in Fig. 7. The most
important difference between
the AM9511 and the MM57109
is that the AM9511 is a binary
device. All data operations
within the AM9511 handle
binary data; in contrast, the
MM57109 handles only BCD
data. AM9511 data may be
specified in fixed-point or float-
ing-point format. Fixed-point
numbers may be single- or
double-precision; in each case
they are treated as signed
binary numbers. A single-pre-
cision fixed-point number is il-
lustrated in Fig. 8.
This is standard signed
binary data. Thus, single-preci-
sion fixed-point numbers may
range in value from - 32768 to
+ 32767. Double-precision
fixed-point numbers are 32 bits
wide, and again use standard
signed binary data format.
Thus, a double-precision
number may have values in the
range -2147483648 through
+ 2147483647.
Floating-point numbers are
all 32 bits wide, and are inter-
OUT* 583
AIN 284
SF1 163
PF1 185
SF2 163
PF2 185
PRW1 130
PRW2 130
TOGM 157
SMDC* 163
INV 166
for each digit to be input. The high order address
for the number to be Input would typically come
from the second instruction word. The digit is in-
put on D4-D1, using ISEL = 0 to select digit data
instead of instructions. The number of digits to be
input depends on the calculation mode (scientific
notation or floating point) and the mantissa digit
count (see Data Formats and Instruction Timing).
Data to be Input is stored In X and the stack is
pushed (X -*• Y -* Z -* T). At the conclusion of the
input, DA4-DA1 = 0.
Addressing and number of digits is identical to IN
instruction. Each time a new digit address is sup-
plied, the processor places thejligit to be output
on D04-D01 and pulses the R/W line active low.
At the conclusion of output, D04-D01 = 0 and
DA4-DA1 = 0.
A single digit is read into the processor on D4-D1.
ISEL = 0 is used by external hardware to select
the digit i nstea d of instruction. It will not read the
digit until ADR = 0(ISEL = 0 selects ADR instead
of I5), indicating data valid. F2 is pulsed active low
to acknowledge data just read.
Set FI high, i.e., FI = 1.
FI is pulsed active high. If FI is already high, this
results in it being set low.
Set F2 high, i.e., F2 = 1.
F2 is pulsed active high. If F2 is already high, this
results in it being set low.
Generates RJ\N active low pulse which may be
used as a strobe or to clock extra instruction bits
into a flip-flop or register.
Identical to PRW1 instruction. Advantage may be
taken of the fact that the last 2 bits of the PRW1
op code are 10 and the last 2 bits of the PRW2 op
code are 01. Either of these^bits can be clocked in-
to a flip-flop using the R/W pulse.
Change mode from floating point to scientific
notation or vice versa, depending on present
mode. The mode affects only the IN and OUT in-
structions. Internal calculations are always in
8-digit scientific notation.
Mantissa digit count is set to the contents of the
second instruction word ( = 1 to 8).
Set inverse mode for trig or memory function In-
struction that will immediately follow. Inverse
mode is for next instruction only.
preted as in Fig. 9. The man-
tissa and exponent are both
binary numbers; therefore,
numbers in the range ±(2.7 x
10' 20 to 9.2 x 10‘ 18 ) may be
represented.
Observe that the AM951 1 has
a smaller range of valid
numbers than the MM57109.
You might argue that the
AM9511, by handling numbers
in the exponential range 10’20
through 10 18 , must surely have
a range adequate for any ap-
plication. This is not always
true.
In particular, chemical-en-
gineering and astronomical
computations frequently han-
dle numbers outside the range
allowed by the AM9511. The
principal advantage of the
AM9511 over the MM57109 is
that the former is much faster.
Table 2 summarizes AM9511 in-
structions. Notice that the in-
struction sets for the two
devices are approximately
CHIP SELECT
CONTROL /DATA
Al
AO
READ STROBE
WHITE STROBE
Fig. 5.
Fig. 6.
Fig. 7. AM9511 arithmetic processor functional logic.
87
H 1 I II 1 1 II 1 11 II M BIT NUMBER
NUMBER
SIGN
Fig. 8. A single-precision fixed-point number.
rr i r~H”i i i i n i i i i i i i i i i i i i
E
MANTISSA
EXPONENT
EXPONENT SIGN
MANTISSA SIGN
Fig. 9.
CS C/D RD WR Function
1 X X X Device not selected
0 0 0 1 Read data from device
0 0 10 Write data to device
0 10 1 Read status from device
0 110 Write command to device
Table 3.
equivalent; however, based on
a 500-nanosecond clock, for
the AM9511 it is more than 100
times faster than for the
MM57109. Also, the AM9511 is
incredibly easy to incorporate
into almost any microcomp>'ter
system. Control signals, data
buses and address buses are
typical of an 8080A support
device. The AM9511 is selected
via the chip-select (CS) and C/D
inputs. This is the standard
method used in any 8080A sup-
port device to access data con-
trol and status locations as two
memory addresses or I/O ports.
The standard read and write
control strobes are used to in-
put or output data. Thus, the
CS, C/D, RD and WR controls
together identify events as in
Table 3.
Data and instructions are in-
put via the bidirectional data
bus; results and status are out-
put via the same bus. While the
AM9511 is busy executing any
operation, a PAUSE signal is
output low. At the end of the
operation the END control sig-
nal is output low. The micropro-
cessor acknowledges the END
output by inputting EACK low.
Any command output to the
AM9511 can, in addition to all
other options, specify a service
request to follow completion of
the AM951 1 operation. During a
service request, CPU will pro-
cess AM9511 results before ini-
tiating a new AM9511 opera-
tion. If a service request is
specified, when the AM9511
completes any operation it out-
puts a low service-request sig-
nal. The CPU acknowledges
this signal with a service-ac-
knowledge input. Thus, the
AM9511 allows the micropro-
cessor to differentiate between
an AM9511 operation that does
or does not require further
handling by the CPU.
When you compare the
AM9511 and MM57109 devices,
selection should be based on
the following trade-offs:
1. The MM57109 is a BCD de-
vice and will therefore be easier
to use in a purely decimal ap-
plication.
2. The MM57109 has a larger
numeric range; however, you
should be sure that the exten-
sive AM9511 numeric range is
insufficient before you go to
the MM57109 based upon this
criterion.
3. The AM9511 is significant-
ly faster than the MM57109.
There may be applications in
which the AM9511 must be
selected based on its speed,
even if BCD-to-binary and
binary-to-BCD conversions are
required.
4. The AM9511 fits naturally
into any 8080A microcomputer
configuration; its bus and con-
trol signal interface is abso-
lutely compatible with the
8080A. In contrast, the
MM57109 is a calculator part
that will need multiplexing and
de-multiplexing circuits sur-
rounding it.
Whether you choose the
AM9511 or the MM57109, you
will be making the right choice
if your alternative is to write
your own transcendental-
function calculations. ■
COMMAND
CLOCK
MNEMONIC
CYCLES
COMMAND DESCRIPTION (1)
SADD
17
Adds TOS to NOS. Result to NOS. Pop Stack
SSUB
30
Subtracts TOS from NOS. Result to NOS. Pop Stack
SMUL
92
Multiplies NOS by TOS. Result to NOS. Pop Stack
SDIV
92
Divides NOS by TOS. Result to NOS. Pop Stack
DADD
21
Adds TOS to NOS. Result to NOS. Pop Stack
DSUB
38
Subtracts TOS from NOS. Result to NOS. Pop Stack
DMUL
208
Multiplies NOS by TOS. Result to NOS. Pop Stack
DDIV
208
Divides NOS by TOS. Result to NOS. Pop Stack
FADD
56-350
Adds TOS to NOS. Result to NOS. Pop Stack
FSUB
58-352
Subtracts TOS from NOS. Result to NOS. Pop Stack
FMUL
168
Multiplies NOS by TOS. Result to NOS. Pop Stack
FDIV
171
Divides NOS by TOS. Result to NOS. Pop Stack
SORT
800
Square Root of TOS. Result in TOS.
SIN
4464
Sine of TOS. Result in TOS.
COS
4118
Cosine of TOS. Result In TOS.
TAN
5754
Tangent of TOS. Result in TOS.
ASIN
7668
Inverse Sine of TOS. Result in TOS.
ACOS
7734
Inverse Cosine of TOS. Result In TOS.
ATAN
6006
Inverse Tangent of TOS. Result in TOS.
LOG
4490
Common Logarithm (base 10) or TOS. Result in TOS.
LN
4478
Natural Logarithm (base e) of TOS. Result in TOS. *>
EXP
4616
Exponential (e x ) of TOS. Result in TOS.
PWR
9292
NOS raised to the power in TOS. Result to NOS. Pop Stack.
NOP
4
No Operation
FIXS
92-216
Converts TOS from floating-point to single-precision fixed-point format.
FIXD
100-346
Converts TOS from floating-point to double-precision fixed-point format.
FLTS
98-186
Converts TOS from single-precision fixed-point to floating-point format.
FLTD
98-378
Converts TOS from double-precision fixed-point to floating-point format.
CHSS
26
Changes sign of single-precision fixed-point operand on TOS.
CHSD
34
Changes sign of double-precision fixed-point operand on TOS.
CHSF
16
Changes sign of floating-point operand on TOS.
PTOS
16
Push single-precision fixed-point operand on TOS to NOS.
PTOD
20
Push double-precision fixed-point operand on TOS to NOS.
PTOF
20
Push floating-point operand on TOS to NOS.
POPS
10
Pop single-precision fixed-point operand from TOS. NOS becomes TOS.
POPD
12
Pop double-precision fixed-point operand from TOS. NOS becomes TOS.
POPF
12
Pop floating-point operand from TOS. NOS becomes TOS.
XCHS
18
Exchange single-precision fixed-point operands TOS and NOS.
XCHD
26
Exchange double-precision fixed-point operands TOS and NOS.
XCHF
26
Exchange floating-point operands TOS and NOS.
PUPI
16
Push floating-point constant “n” onto TOS. Previous TOS becomes NOS.
Notes: 1. Nomenclature: TOS is Top Of Stack. NOS is Next On Stack.
2. All derived floating-point functions destroy the contents of the stack. Only the result can be
counted on the be valid upon command completion.
3. Format conversion commands (FIXS, FIXD, FLTS, FLTD) require that floating-point data format be
specified (command bits 5 and 6 must be 0).
Table 2. AM9511 instruction description table.
88
ELECTRONIC SYSTEMS
P.O. Box 9641 San Jose C A 95157
(408) 374-5984
UART
& BAUD
RATE
GENERATOR
Part no. 101
• Converts serial to parallel and
parallel to serial
• Low cost on board baud rate
generator
• Baud rates: 110, 150,
300, 600, 1200, and 2400
• Low power drain +5 volts and
-12 volts required
• TTL compatible
• All characters contain a start
bit, 5 to 8 data bits, 1 or 2 stop
bits, and either odd or even
parity.
• All connections go to a 44 pin
gold plated edge connector
• Board only $12.00; with parts
$35.00
8K
STATIC
RAM
Part no. 300
• 8K Altair bus memory
• Uses 2102 Static memory chips
• Memory protect
• Gold contacts
• Wait states
• On board regulator
• S-100 bus compatible
• Vector input option
• TR1 state buffered
• Board only $22.50; with parts
L $160.00
RS* 232 / TTL
INTERFACE
Part no. 232
• Converts TTL to RS-232, and
converts RS-232 to TTL
• Two separate circuits
• Requires -12 and +12 volts
• All connections go to a 10 pin
gold plated edge connector
• Board only $4.50; with parts
$7.00
DC
POWER
SUPPLY
Part no. 6085
• Board supplies a regulated +5
volts at 3 amps., +12,-12, and -5
volts at 1 amp.
• Circuit has filters, rectifiers,
and regulators.
• Power required is 8 volts AC at
3 amps., and 24 volts AC C.T. at
1.5 amps.
• Board only $12.50
TIDMA
Part no. 112
• Tape Interface Direct Memory
Access
• Record and play programs with-
out bootstrap loader (no prom)
has FSK encoder/decoder for
direct connections to low cost
recorder at 625 baud rate, and
direct connections for inputs and
outputs to a digital recorder at
any baud rate.
• S-100 bus compatible
• Board only $35.00;
with parts $110.00
Part no. Ill
TAPE
INTERFACE
• Play and record Kansas City
Standard tapes
• Converts a low cost tape
recorder to a digital recorder
• Works up to 1 200 baud
• Digital in and out are TTL-serial
• Output of board connects to
mic. in of recorder
• Earphone of recorder connects
to input on board
• Requires +5 volts, low power
drain
• Board $7.60; with parts $27.50
• No coils
Part
no. 107
RF
MODULATOR
• Converts video to AM modu-
lated RF, Channels 2 or 3
• Power required is 12 volts AC
C.T., or +5 volts DC
• Board $7.60; with parts $13.50
■m
Apple II
Serial I'D
Interface —
Part No. 2
• Baud rates up to 30,000
• Plugs into Apple Peripheral
connector
• Low-current drain
• RS-232 Input and Output
SOFTWARE
• Input and Output routine from
monitor or BASIC to teletype or
other serial printer.
• Program for using an Apple II
for a video or an intelligent ter-
minal. Board only — $15.00;
with parts — $42.00; assembled
and tested — $62.00.
TELEVISION
TYPEWRITER
a Li ^ j
!
easj
Part no. 106
• Stand alone TVT
• 32 char/line, 16 lines, modifi-
cations for 64 char /line included
• Parallel ASCII (TTL) input
• Video output
• 1 K on board memory
• Output for computer con-
trolled curser
• Auto scroll
• Non-distructivc curser
• Curser inputs: up, down, left,
right, home, EOL, EOS
• Scroll up, down
• Requires +5 volts at 1.5 amps,
and -12 volts at 30 mA
• Board only $39.00; with parts
$145.00
MODEM
Part no. 109
• Type 103
• Full or half duplex
• Works up to 300 baud
• Originate or Answer
• No coils, only low cost com-
ponents
• TTL input and output-serial
• Connect 8 ohm speaker and
crystal mic. directly to board
• Uses XR FSK demodulator
• Requires +5 volts
• Board $7.60; with parts $27.50
To Order:
Mention part number and description. For parts kits add “A” to part number. Shipping paid for orders
accompanied by check, money order, or Master Charge, BankAmericard, or VISA number, expiration
date and signature. Shipping charges added to C.O.D. orders. California residents add 6.5% for tax.
Parts kits include sockets for all ICs, components, and circuit board. Documentation is included with
all products. Dealer inquiries invited. 24 Hour Order Line: (408) 374-5984. £21
Ken Barbier
Borrego Engineering
PO Box 1253
Borrego Springs CA 92004
Money Manipulations
keep ahead of
those cash-flow problems
I REM CASH FLOW PROGRAM 29 JULY 77
7A$= <.« $ ######. ##»
8 B $=“####”
9 PRINT: PRINT: PRINT
10 PRINT" IN VESTMENT MINUS DRAW”
II PRINT" QUARTERLY STATEMENT”
15 PRINT
20 PRINT‘‘PRINCIPAL: $; INTEREST: %/YR; DRAW: $/MO.”
21 PRINT" INTEREST EARNED SHOWN BY QUARTER TOTAL”
22 PRINT" DRAW IS CURRENT MONTHLY RATE, INFLATED”
25 PRINT:PRINT
30 M=0
31 Y=0
35 C=2
40 INPUT "PRINCIPAL=”;P
50 INPUT "INTEREST=”;I
60 INPUT “DRAW=”;D
61 INPUT "INFLATION=”;A
62 A=A/1200
70 1=1/1200
80 PRINT: PRINT: PRINT
90 PRINT"MONTH PRINCIPAL EARNED DRAW”
100 E=P*I
110 P=P+E
120 P=P-D
125 IF P<0 THEN 190
130 M=M+1
135 D=D*(1 + A)
137 Y=Y+E
138 IF M=123 THEN 180
140 IF M >240 THEN 1200
141 IF C=0 GOTO 144
142 C=C-1
143 GOTO 100
144 C=2
146 PRINT USING B$;M;:PRINT‘‘ ”;
150 PRINT USING A$;P,Y,D
151 Y=0
170 GOTO 100
180 PRINT: PRINT: PRINT: PRINT
181 GOTO 141
190 PRINT :PRINT:PRINT: PRINT :PRINT
1200 END
Fig. 1. Cash Flow program listing. I/Vritten in 12K Extended BASIC,
it can be run on smaller BASICS by changing lines 146 and 150 to:
150 PRINT M,P,Y,D. Lines 7 and 8 can then be dropped. The
output will not be so nicely formatted however.
Y ou say you're getting
ready to punch your
boss's lights out, but you're
not sure your life savings will
support the wife and kiddies
until you get out of jail? Or
maybe you're just getting old
(like me) and think it's time
to retire, but you want to be
sure you have enough loot
stashed away to supplement
Uncle Sam's pittance and pro-
vide enough to live on —
forever and ever. Or perhaps
you are ready to throw in the
towel at the boiler factory
and open your own computer
store . . . and want to know
how long you can hold out
until the first cash customer
comes walking in. Well, tell
you what I'm going to do . . .
Computing Cash Flow
The Cash Flow program
listing in Fig. 1 assumes that
an initial investment is made
at a fixed rate of interest
(compounded monthly). But
instead of simply figuring
compound interest. Cash
Flow assumes that we will be
drawing on these reserves, for
reasons such as those listed
above. Furthermore, life be-
ing what it is, the amount we
have to withdraw will be sub-
ject to inflation, so the pro-
gram takes this factor into
account as well. Since Uncle
Sam insists we pay income
tax on the interest paid on
our investments, we will also
need a statement showing in-
terest earned. While the pro-
gram will not fill out your
income tax form for you, it
will, considering all these
factors, tell you how long
your loot will last.
For example, let's take a
look at a typical Cash Flow
run (Fig. 2). Dick and Jane
have both been working and
diligently squirreling money
away. They have accumulated
forty kilobucks and would
like to use it to finance an
early retirement. What they
need to know is whether or
not the money will hold out
until social security helps
them out (assuming it doesn't
go broke first).
Being conservative, they
will invest the money in in-
90
sured savings, which, for our
example, we will assume pays
5.75 percent per year, com-
pounded monthly. They have
moved into a less expensive
house, but there are still pay-
ments to make. Now, our
couple must figure the maxi-
mum amount per month that
they will have to draw from
their savings to live on. This
fictional account shows that
they have arrived at a figure
of $750 per month, which
certainly should be enough to
feed two mouths.
Next, we throw in a little
magic. D. and J. have con-
sulted their financial expert,
and he assures them they can
expect an inflation rate of 3
percent per annum to apply
to the commodities they will
be consuming. This figure
sounds low today, but if
coffee, new cars, etc., are
avoided it is not too unrealis-
tic.
All the above conditions
established, we load Cash
Flow, which is written in
Altair BASIC, 12K Extended,
Version 3.2. Instructing it to
run, we are informed that we
will be provided with a quar-
terly statement, and we are
asked to enter the amount of
principal (in dollars); the
interest rate (in percent per
annum); the amount we wish
to withdraw (in dollars per
month); and the expected
annual rate of inflation. Hav-
ing received these variables,
Cash Flow proceeds to pro-
duce the quarterly-statement
table shown in Fig. 2.
Since this is a quarterly
statement, the number of the
month for which the figures
apply will increment by
three. The amount of princi-
pal remaining at the end of
that month is shown in the
next column. The third col-
umn shows the total interest
earned for the previous
quarter, which is what we will
have to pay income tax on.
This last column shows our
draw for the current month.
This amount always increases
because we have to assume
that inflation will continue to
spiral.
When all of the money is
used up. Cash Flow will ter-
minate, and we will have to
go back to work. We see that
Dick and Jane can survive for
about five years. Well, maybe
they'd better try to cut costs
a little. Then we can try the
program again, using a lower
Draw figure.
When this program was
first run, the nice round num-
bers in the cents column
under Principal raised sus-
picion. The BASIC manual
states that single-precision
numbers are printed with a
maximum of six decimal dig-
its, and we are asking BASIC
to work with seven digits! So,
we should add the following
line to our program: 2
DEFDBL P.
Now when we run the
program with the same vari-
ables, we get the output
shown in Fig. 3, since Princi-
pal is computed in double
precision. We can see the
pennies and nickles, but the
results don't change! This is
because we had sufficient ac-
curacy to begin with, the
internal representation of our
principal being in binary bits,
which don't exactly relate
evenly to six-decimal digits.
Our initial accuracy was bare-
ly sufficient, though, so it
would be a good idea to leave
the second line in our pro-
gram, in case a rich uncle dies
and leaves more money to
play with.
Since Dick and Jane are
only 23 years old (surprise!)
they have decided to post-
pone the early retirement and
keep on working and saving.
Now they can use the same
program to estimate how
their savings will grow if left
untouched. If no money is
drawn from the investment,
Cash Flow becomes a
straightforward compound-
interest program, as we can
see in Fig. 4.
Here, we set draw and
inflation to zero, and Cash
Flow gives a quarterly state-
ment of earnings and accumu-
lation for our savings ac-
count. The program gets tired
and quits after 20 years. Dick
and Jane probably will,
too! ■
INVESTMENT MINUS DRAW
QUARTERLY STATEMENT
PRINCIPAL: $; INTEREST: %/YR; DRAW: $/MO.
INTEREST EARNED SHOWN BY QUARTER TOTAL
DRAW IS CURRENT MONTHLY RATE, INFLATED
PRINCIPAL=? 40000
INTEREST=? 5.75
DRAW=? 750
INFLATION? 3
MONTH
PRINCIPAL
EARNED
DRAW
3
$38311.30
$566.95
$755.64
6
$36581.20
$542.48
$761.32
9
$34808.90
$517.41
$767.05
12
$32993.80
$491.73
$772.81
15
$31135.00
$465.43
$778.62
18
$29231.70
$438.49
$784.48
21
$27283.30
$410.92
$790.38
24
$25289.00
$382.69
$796.32
27
$23247.80
$353.80
$802.31
30
$21159.10
$324.23
$808.34
33
$19022.00
$293.97
$814.42
36
$16835.70
$263.02
$820.54
39
$14599.20
$231.35
$826.71
42
$12311.90
$198.96
$832.93
45
$9972.66
$165.83
$839.19
48
$7580.74
$131.95
$845.50
51
$5135.21
$97.31
$851.86
54
$2635.15
$61.90
$858.26
57
$79.62
$25.69
$864.71
Fig. 2. Sample Cash Flow run. This printout shows how long an
initial investment of $40,000 will last while earning 5.75 percent
interest, but being drawn on at the rate of $750 per month, inflated
3 percent per year.
INVESTMENT MINUS DRAW
QUARTERLY STATEMENT
PRINCIPAL: $; INTEREST: %/YR; DRAW: $/MO.
INTEREST EARNED SHOWN BY QUARTER TOTAL
DRAW IS CURRENT MONTHLY RATE, INFLATED
PRINCIPAL=? 40000
INTEREST=? 5.75
DRAW=? 750
INFLATION=? 3
MONTH
PRINCIPAL
EARNED
DRAW
3
$38311.32
$566.95
$755.64
6
$36581.21
$542.48
$761.32
9
$34808.94
$517.41
$767.05
12
$32993.78
$491.73
$772.81
15
$31134.96
$465.43
$778.62
18
$29231.74
$438.49
$784.48
21
$27283.34
$410.92
$790.38
24
$25288.97
$382.69
$796.32
27
$23247.83
$353.80
$802.31
30
$21159.12
$324.23
$808.34
33
$19022.01
$293.97
$814.42
36
$16835.66
$263.02
$820.54
39
$14599.24
$231.35
$826.71
42
$12311.86
$198.96
$832.93
45
$9972.66
$165.83
$839.19
48
$7580.75
$131.95
$845.50
51
$5135.22
$97.31
$851.86
54
$2635.15
$61.90
$858.26
57
$79.62
$25.69
$864.71
Fig. 3. A double-precision run. The net results have not changed,
but would for larger principals. Double precision results in a more
accurate printout, but the program takes longer to run.
INVESTMENT MINUS DRAW
QUARTERLY STATEMENT
PRINCIPAL: $; INTEREST: %/YR; DRAW: $/MO.
INTEREST EARNED SHOWN BY QUARTER TOTAL
DRAW IS CURRENT MONTHLY RATE, INFLATED
PRINCIPAL=? 10000
INTEREST=? 6
91
<^S>
Can your computer
read and solve
this problem by itself?
“ ON THEIR VACATION, TOM AND
DICK VISITED A FARM. WHILE
THERE, THEY NOTICED A PEN
CONTAINING CHICKENS AND
PIGS. TOM SAID THERE WERE 3
TIMES AS MANY CHICKENS AS
PIGS. DICK SAID HE COUNTED
100 LEGS IN THE PEN. HOW
MANY CHICKENS WERE IN THE
PEN?”
with NLOS/l, it can!
NLOS/1 is a cassette-based
/ n. system requiring a minimum
( ) ) of 12K, a serial I/O board and
J any cassette interface. The
I system comes complete with
v/J a fully documented set of as-
^ sembly language source list-
ings. The cost is only $50.00.
STOP
PROGRAMMING
YOUR COMPUTER,
EDUCATE IT !
ORDER TODAY!
CYBERMATE C57
R.D. #3 BOX 192A
NAZARETH PA 18064
DRAW =? 0
INFLATION =? 0
MONTH
PRINCIPAL
EARNED
DRAW
3
$10150.75
$150.75
$0.00
6
$10303.78
$153.02
$0.00
9
$10459.11
$155.33
$0.00
12
$10616.78
$157.67
$0.00
15
$10776.83
$160.05
$0.00
18
$10939.29
$162.46
$0.00
21
$11104.20
$164.91
$0.00
24
$11271.60
$167.40
$0.00
27
$11441.52
$169.92
$0.00
30
$11614.00
$172.48
$0.00
33
$11789.08
$175.08
$0.00
36
$11966.81
$177.72
$0.00
39
$12147.21
$180.40
$0.00
42
$12330.33
$183.12
$0.00
45
$12516.21
$185.88
$0.00
48
$12704.89
$188.68
$0.00
51
$12896.42
$191.53
$0.00
54
$13090.83
$194.42
$0.00
57
$13288.18
$197.35
$0.00
60
$13488.50
$200.32
$0.00
63
$13691.84
$203.34
$0.00
66
$13898.25
$206.41
$0.00
69
$14107.77
$209.52
$0.00
72
$14320.44
$212.68
$0.00
75
$14536.33
$215.88
$0.00
78
$14755.46
$219.14
$0.00
81
$14977.90
$222.44
$0.00
84
$15203.70
$225.79
$0.00
87
$15432.89
$229.20
$0.00
90
$15665.55
$232.65
$0.00
93
$15901.71
$236.16
$0.00
96
$16141.43
$239.72
$0.00
99
$16384.76
$243.33
$0.00
102
$16631.76
$247.00
$0.00
105
$16882.49
$250.73
$0.00
108
$17136.99
$254.51
$0.00
111
$17395.34
$258.34
$0.00
114
$17657.57
$262.24
$0.00
117
$17923.76
$266.19
$0.00
120
$18193.97
$270.20
$0.00
123
$18468.24
$274.28
$0.00
126
$18746.65
$278.41
$0.00
129
$19029.26
$282.61
$0.00
132
$19316.13
$286.87
$0.00
135
$19607.32
$291.19
$0.00
138
$19902.91
$295.58
$0.00
141
$20202.95
$300.04
$0.00
144
$20507.51
$304.56
$0.00
147
$20816.66
$309.15
$0.00
150
$21130.47
$313.81
$0.00
153
$21449.02
$318.54
$0.00
156
$21772.37
$323.35
$0.00
159
$22100.59
$328.22
$0.00
162
$22433.76
$333.17
$0.00
165
$22771.95
$338.19
$0.00
168
$23115.24
$343.29
$0.00
171
$23463.70
$348.47
$0.00
174
$23817.42
$353.72
$0.00
177
$24176.47
$359.05
$0.00
180
$24540.94
$364.46
$0.00
183
$24910.89
$369.96
$0.00
186
$25286.43
$375.53
$0.00
189
$25667.62
$381.20
$0.00
192
$26054.57
$386.94
$0.00
195
$26447.34
$392.78
$0.00
198
$26846.04
$398.70
$0.00
201
$27250.75
$404.71
$0.00
204
$27661.55
$410.81
$0.00
207
$28078.56
$417.00
$0.00
210
$28501.84
$423.29
$0.00
213
$28931.51
$429.67
$0.00
216
$29367.66
$436.15
$0.00
219
$29810.38
$442.72
$0.00
222
$30259.78
$449.40
$0.00
225
$30715.95
$456.17
$0.00
228
$31178.99
$463.05
$0.00
231
$31649.02
$470.03
$0.00
234
$32126.13
$477.11
$0.00
237
$32610.44
$484.31
$0.00
240
$33102.04
$491.61
$0.00
Fig. 4. Compound-interest run. If Draw is set to zero. Cash Flow
becomes a straight compound-interest computation. Here, $10,000 was
invested at 6 percent for 20 years. Changing program line 140 can vary
this time limit.
92
COME ON UP AND
SEE ME SOMETIME
™mo©[^§>
Give your 6800 computer the gift of sight! The Micro Works Digisector" opens up a whole new
world for your computer. Your micro can now be a part of the action, taking pictures like this
one to amuse your friends, watching your home while you're away, helping your household
robot avoid bumping into walls, providing fast to slow scan conversion for you hams . the
applications abound.
The Micro Works Digisector is a completely unique device; its resolution and speed are un-
matched in industry and the price is unbeatable anywhere. The Digisector and a cheap TV
camera are all you'll need to see eye to eye with your 6800. Since operation is straightforward,
you don't have to be a software wizard to utilize the Digisector's extensive capabilities. The
Micro Works Digisector board provides the following exclusive features:
• High Resolution— a 256 x 256 picture element scan
• Precision— 64 levels of grey scale
• Speed— Conversion times as low as 3 microseconds per pixel
• Versatility— Accepts either interlaced (NTSC) or non-interlaced (Industrial) video input
• Compactness— Utilizes 1 I /O slot in your SWTPC 6800 or equivalent
• Economy— The Digisector is a professional tool priced for the hobbyist
The Digisector (DS-68), like all Micro Works products, comes fully assembled, tested and
burned in. Only the highest quality components are used, and the boards are double sided
with plated through holes, solder mask and silkscreen. All software is fully source listed and
commented. The Micro Works is proud to add the DS-68 to its line of quality computer acces-
sories for the hobbyist. Price 169.95
29 95
9.95
24.95
29.95
22 95
Visa and Master Charge Accepted
P.O.BOX 111Q DEL MAR, CA. 92014 714-756-2687 M31
The Micro Works 6800 series of computer accessories also includes.
PSB-08 PROM System Board 11 9.95
regulated + 1 2 volts 124.95
B-08 2708 EPROM Programmer 99.95
regulated + 1 2 volts 104.95
U2708 EPROM Software
Casettetape
UIO Universal I/O Board
X-50 Extender Board
X-30 Extender Board
RAM, PROM, EPROM, PARALLEL I/O, SERIAL I/O, A/D, D/A, CPU
Watch For New Product Releases
25 SOUTH 300 EAST SUITE 215* SALT LAKE CITY, UTAH 84111 • 714/752-1374
93
Strings and Things
BASIC conversion techniques
Richard Roth
TSA Software
5 N. Salem Road
Ridgefield CT 06877
Y ou have advanced far
enough in programming to
use character strings; yet,
when you try to run a program
using character strings from a
book or article, you find half of
them don’t make any sense. If
so, or if you are interested in
handling characters in general,
this article is for you.
A character string, basically
a one-dimensional array or vec-
tor of characters, is a sequence
of characters one after
another. What distinguishes it
from a vector of numbers is that
it is used as a whole, rather
than a character at a time.
In a game, the program may
ask for someone’s name, but it
doesn’t care that JOHN is a J
followed by an O, H, N. The indi-
vidual letters are considered a
unit. In contrast, a mailing-list
program that prints a list by
last names scans MARYbb J.b-
JONES to find the last word. It
does this by scanning the char-
acters until it finds a sequence
of characters followed only by
blanks. A space (represented
by b or blank) breaks the se-
quence of characters that com-
prise a word. We call such a
break character a delimiter.
Commas and periods also
break the sequences of words
into smaller units— phrases
and sentences. A smaller unit
of a character string is called a
substring. Another special fea-
ture of character strings is
length; a unit called NAME can
vary from ED to STASTICOVICH.
Usually, we fix a maximum
length, but often we want to
know the current length.
The problem arises when you
want to use strings in BASIC,
originally intended to work with
numbers. Of course, a letter
can be represented by a num-
ber, such as A = 1, B = 2 ... or
by the ASCII character set. In
ASCII, digits (0-9), letters and
special characters (such as
Bell or Return) are all repre-
sented by a single integer from
0 through 127 (funny— it just
fits in one byte!). In working
with such simple numbers,
BASIC wastes space because it
is prepared for many digits of
precision and doesn’t know
how simple a number is. Deal-
ing with varying length and the
string as a unit requires some
built-in features. In the scienti-
fic language FORTRAN, the
programmer must have a whole
set of special subroutines to
deal with strings.
When BASIC was first devel-
oped by Dartmouth’s Kemeny
and Kurtz, the only strings
allowed were literals in print
statements for title and labels
such as: 100 PRINT “X = ”,X. In
early versions of BASIC, such
as GE-635 Mark I Timesharing,
extensions were added to allow
the storage of strings, which
were handled like single num-
bers. However, no advanced
capability was available. A
string array was specified by
giving it a two-letter name. All
one could do was print the
DIM AA(3)
READ AA(1), AA(2)
, AA(3)
DATA ‘SALLY’, ‘JOE’, ‘SPOT’
PRINT ‘A GIRL IS’,
, AA(1)
PRINT ‘A BOY IS’,
AA(2)
PRINT ‘A DOG IS’,
AA(3)
Example 1.
HP BASIC
DEC BASIC
HP
DEC
Data General (DG)
Mits/Microsoft
North Star
BASIC-E
Computer Science Corp.
Tymshare
Micro-polius (??)
Table 1.
94
HP
100 DIM N$(30),L$(10)
110 S^O (state is beginning)
120 C = 1 (character 1)
130 N$ = “SALLYbbJ.b JONES”
140 IF C>LEN(N$) GO TO 200
1 50 IF N$(C,C) = “b ’ THEN S = C
160 C = C+1
170 GO TO 140
200 REM Now S = Char of last space
210 L$ = N$(S+ 1)
DEC
100 REM
110 S = 0
120 C = 1
130 N$ = “SALLYbbJ.b JONES”
140 IF C>LEN(N$) GO TO 200
150 IF MID$(N$,C,1) = “b”THEN S = C
160 C = C+1
170 GO TO 140
200 REM
210 L$ = RIGHT$(L$,S + 1)
Example 2 .
specific character. By HP rules
the first subscript is the start-
ing character, the second is
ending character: if A$ =
“ABCD”, then A$ (2,3) = “BC”.
If no second subscript exists,
then the rest of the string is
used: if A$ = “ABCDEF” then
A$(3) = “CDEF”. DEC uses
functions MID$, RIGHTS,
LEFTS (as shown in the table)
for the string AS = “ABCDEF”.
Single-character string func-
tions are shown in Table 4.
whole string. (See Example 1.)
This extension was short
lived, but it set the stage for
what we now have. The idea of
uniquely specifying a string
name became more prevalent,
and ‘S’ was finally accepted as
the last character of a string
name. But we are still plagued
by the questions: How does
one specify a single character
of a string; and how does one
specify a matrix of strings?
Two primary approaches devel-
oped— one by Digital Equip-
ment Corporation (DEC) and
the other by Hewlett-Packard
(HP) in the HP-200 series.
DEC emphasized many
strings grouped as a matrix;
and so A$(1) became the first
element in a string matrix. HP
emphasized each character in
the string; and so A$(1) became
the first character of string AS.
These approaches led to a ma-
jor difference in string han-
dling. DEC BASIC requires a
special way of getting a single
character of a string, while HP
BASIC must handle a string ar-
ray specially. Table 1 shows a
summary of the different
BASICS.
From now on I will refer to the
two schemes simply as HP or
DEC, even though most
schemes I will be referring to
have not been written by either
company.
What’s It All Mean?
The issue involves how one
deals with strings. For simple
strings, such as printing the
name of a single game-player,
there is (almost) no difference
(see Table 2).
Since HP BASIC uses the
subscript notations to refer to
substrings, DIM specifies the
length of the string. DEC BASIC
uses DIM to indicate how many
strings in a string matrix; if no
DIM occurs, it is just a single
string (also called a scalar
string, as opposed to a matrix).
DEC BASIC has no way of
specifying maximum string
length. They allow a maximum
limit, usually 255 characters,
set by the BASIC designer. HP
BASIC tends to allow length
limited only by memory size.
Since HP BASIC knows how big
a string can get, it can reserve a
fixed space. DEC BASIC must
constantly shift the strings
around as lengths change. In
this respect, HP BASIC enjoys
a speed advantage.
Character Manipulations
There are two levels of ma-
nipulations— character and
string. Each scheme of BASIC
has its own home ground:
character for HP and string for
DEC. Getting at a single
character is required for many
functions. An early example
suggested extracting a last
name to alphabetize a mailing
list. Example 2 shows this in
both schemes.
To get at the fifth character
of a string, HP BASIC uses only
N$(5,5), while DEC BASIC uses
MID$(N$,5,1). In our example,
the program considers one
character at a time from the
string N$ with the name, until it
gets to the end. Each time it
sees a blank, it saves the char-
acter number in S. With no trail-
ing blanks, the program, when
it reaches line 200, S will point
to the last blank. So the last
name is S + 1 through the end.
(For simplicity, I am assuming a
statement can follow an IF-
THEN statement, as in most
current BASICS.)
Table 3 shows how to get at a
HP
DIM A$(30)
INPUT AS
PRINT “HELLO”, AS
RUN
?SAM (CR)
HELLO SAM
String Manipulations
While DEC BASIC has awk-
ward functions when dealing
with substrings, HP BASIC has
a far greater problem when
many strings must be manipu-
lated. It has no way to handle a
group of strings of variable
length. In the HP-3000 BASIC
this was remedied in an elegant
manner— especially true to
BASIC syntax; unfortunately,
only HP-3000 and Computer
Science have implemented this
INPUT AS
PRINT “HELLO”, AS
RUN
?SAM (CR)
HELLO SAM
Table 2.
DEC HP
MID$(A$, starting char, length)
MID$(A$,2,2)= “BC” A$(2,3)
LEFT$(A$, length)
LEFT$( AS , 3) = “ABC ” A$( 1 ,3)
RIGHT$(A$, starting char)
RIGHT$(A$,3) = “CDEF” A$(3)
(Note: The parameter for RIGHTS is starting
character in DEC BASIC-plus, but length
from the right in Mits BASIC and BASIC-E.)
Table 3.
Function
Length
Substring - 1 char at
- N chars at I
- Char I to J
- Char I to end
- Char 1 to Char I
HP
DEC
LEN(A$) LEN(A$)
A$(I,I) MID$(A$,I,1)
A$(I,I + N -1) MID$(A$,I,N)
A$(I,J) MID$(A$,I,J-I + 1)
A$(I) RIGHT$(A$,I)
A$(1,I) LEFT$(A$,I)
Table 4.
DEC
95
in their BASICS. They have
added:
DIM S(3)$(5)
A$ = S(1)$(3,4)
The first subscript is a matrix;
the second, a substring.
Similarly, the first DIM value is
matrix size; the second is max-
imum length.
Most people with HP BASIC
can handle (with difficulty) a
form of string matrix. Imagine
the string V$, length 100, to be
made of ten substrings, each
ten characters long. The key is
to fill out each string to a full
ten characters; otherwise, the
larger string will have holes.
Creating one of those holes by
putting in a shorter string will
chop off the rest of the larger
string. This also makes all the
pseudomatrix elements a fixed
length, which is annoying but
better than no string matrix.
For example, the fifth string in
string matrix V$ is extracted by
using: S5$ = V$(4*L + 1 ,5*L)
(this is for a matrix starting at
element 1). Two simple user
functions will ease this calcula-
tion (see Example 3).
Concatenation
The second major function in
string manipulation is concate-
nation, i.e., combining two
strings to make one. For exam-
ple, “HEL” + “LO” = “HELLO”
(using DEC concatenation op-
erator). HP has no common,
direct way of doing this. Both
+ and , are allowed in some HP
BASICS as concatenation oper-
ators. If no operator exists, HP
BASIC allows a rather strange
use of the subscript/substring
to do this (see Table 5). At the
LI and L2 calculations, X$ is
kept at full length and need not
be refilled.
When using HP form strings
for pseudostring matrices or
concatenation, one must be
very careful to fill out each
string assignment where the
subscript/substring is on the
left side, i.e., S$(1,4) = A$. An
improper assignment may
chop off the end of the string on
the left. This varies between
HP-style BASICS, for example:
A$(5,9) = BS where LEN(B$)<4
A$(5) = B$
In both cases, the length of A$
might become 5 + LEN(B$).
(Data General had this problem
before Release 3 RDOS BASIC,
whereas, North Star Release 2
does not have the problem.)
Commands, Special Characters,
Numbers and Input/Data
There are several less impor-
tant differences that relate to
assorted areas that vary be-
DEF FNL(X) = (X- 1)*L + 1 /DEF FNH(X) = (X* L)
A$(FNL(X),FNH(X)) references element X where:
X = subscript, L = length and A$ is pseudomatrix.
Example 3.
HP
100 IF A = B THEN PRINT “EQUALS* ’/GO TO 300
110 GO TO 400
DEC
100 IF A = B THEN PRINT “EQUALS”: GO TO 300
110 GO TO 400
Example 4.
(from DEC BASIC-PLUS
A$ = “BCDEFAF”
INSTR(1,A$,“AF”) = 6 (6th char position)
INSTR(1,A$,“ABD”) = 0 (not found)
INSTR(6,A$,“F”) = 7 (start looking at 6th char)
Example 5.
tween both schemes, all ver-
sions. Commands vary from
BASIC to BASIC, for example,
NEW or SCR (scratch), which is
used to clear out an old pro-
gram.
Getting special characters
into and out of strings requires
special care. Normally, a bell,
for example, cannot be entered
into a string. Some BASICS
allow the code to be typed in a
quoted literal. This can cause a
problem because a listing will
not show the character or, even
worse, it will do the function
(for example, turn on the paper-
tape punch). One scheme by
DG allows a special form in
literal <#> in which the number
is the internal form of the
special code. For example, <7>
is an ASCII BEL Code. The more
common version allows a func-
tion, usually CHR$, that con-
verts the numeric value to a
string of the same character
(BELL Code = CHR$(7)). The
reverse function is ASC for
ASCII value, where ASC(“A”)
= 65 (the value of the letter A in
the ASCII code). (Some BASICS
use an ASCII null (true 0 byte) to
indicate the end of a string. So
A$(10) = CHR$(0) will chop off
the string at 9 characters— if
your BASIC does this.)
A similar conversion from in-
VAL(A$) = 0 if A$ = “0.0”. If
your BASIC does not have a
formatted print, these are
useful in doing special output
or input formatting. Read your
manual before trying these
functions; they might not do
what you would expect. De-
pending on the BASIC, the
following sequence could give
a lot of trouble.
10 A= 10
20 A$ = NUM$(A)
30 FS = “FILE” + A$
40 OPEN FILE F$
Some BASICS format a
“NUM$” call exactly likeoutput
and put a space before the nu-
meric string. For example, F$ =
“FILEb 10” — not “FILE10”
Some special functions allow
any string, expression or literal,
while others must be a simple
variable. (The difference be-
tween internal form of a num-
ber and ASCII byte or a charac-
ter string can be confusing for
the novice. 10 is not the same
as “10” and if you are not sure
why, find someone who knows.
For example, a BEL code is an
ASCII 7, not 7.0 or “7”— the dif-
ference depends on the func-
tion required.) Because it is not
clear which is the “obvious
way,” both exist (see Example
4). DEC style says when the IF
ternal to character string form
is often available for numbers,
too. NUM$(A) = “0.0” if A = 0 or
condition is true execute the
rest of the statement; if it is
false, continue on the next line.
DEC
HP
10
A$ = “HEL”
10
DIM X$(80)
20
B$ = “LO”
20
X$ = “ ” (80 blanks)
30
S$ = A$ + B$
30
A$ = “HEL”
40
PRINT “STRING =
”,S$ 40
B$ = “LO”
Run
50
LI = LEN(A$)
STRING = HELLO
60
L2 = LEN(B$)
70
X$(1,L1) = A$
80
X$(L1 + 1,L1+L2) = B$
90
S$ = X$(1,L1 + L2)
100 PRINT “STRING = ”,S$
Run
STRING = HELLO
Table 5.
5-190 — Dimension and Functions
200-299 — Read in names and Data Statements
300-499 — Swap names, last name first
500-699 — Bubble sort alphabetically
700-899 — Print sorted list
Table 6.
96
Most HP BASICS only allow a
line number after THEN. North
Star says if true, execute the
rest; if false, skip only the
THEN clause, not the line. HP-
style BASIC may or may not
print “EQUALS,” but it will
always go to line 300; DEC style
will only go to line 300 if
“EQUALS” is printed; other-
wise it will go to line 400.
Another feature of some
BASICS is a string search,
which locates a substring in a
larger string (see Example 5).
Back to Reality
Let’s condense all this dis-
cussion into one example
which compares a list sort in
HP-style and DEC-style BASIC.
To add character functions we
enter the list first name first
and sort it first name last. Both
are listed in Programs A and B
and have approximately cor-
responding line numbers (see
Table 6).
For the HP-like BASIC, we
used North Star BASIC, which
took 22 seconds from run to
ready; the DEC-like BASIC was
BASIC-E, which took 10 sec-
onds (but it’s a partial com-
piler). Neither time reflects a
great sort but it works and il-
lustrates our discussion here.
(Fig. 1 is a run of the program.)
Peculiarities of the HP-like
version are primarily related to
the pseudomatrix required be-
cause the names functions
FNL and FNH are used to cal-
culate the start and end charac-
READY
LIST
100 REM WRITTEN IN NORTHSTAR BASIC (RELEASE 2)
110 READ N9
120 DIM N$(N9*30), F$(30),F1$(30),F2$(30),A$(30)
130 REM USE FUNCTIONS FOR PSEUDOMATRIX OF STRINGS
140 DEF FNL(X) = (X - 1)*30 + 1 \ DEF FNH(X) = X*30
150 DEF FNA$(A$)
160 IF LEN(A$)> = 30 THEN RETURN A$
170 A$ = AS + “b” \ GOTO 160 \ FNEND
200 REM IN NAMES
205 PRINT “ **** NAMES ♦***” \ PRINT
210 N$ = \ REM CLEAR MATRIX
220 FOR I = 1 TO N9
230 READ F$
235 PRINT F$
240 F$ = F$+ “$’* \ REM MARK END OF NAME FOR REVERSE ROUTINE
250 F$ = FNA$(F$) \ REM FILL NAME TO 30 CHARS
260 N$ = N$ + F$
270 NEXT I
280 REM DATA
282 DATA 10
284 DATA “SALLY JONES”, “SAM SMITH”, “JOE SMITH”, “TIM CAMBELL”, “ED HILL”
286 DATA “STEVE MOODY”, “ROGER HEAD”, “SHIRLEY JONES”, “ISSAC DEAR”, “RICH KING”
300 REM RE-ORDER LAST NAME FIRST
310 FOR N1 = 1 TO N9
320 F$ = N$(FNL(N1),FNH(N1))
330 C=1
335 REM LOOP UNTIL END MARK FOUND
340 IF F$(C,C) = “$” THEN 380
350 IF F$(C,C) = “b” THEN S = C
360 C = C+1
370 GOTO 340
380 REM REVERSE FIRST & LAST NAMES
390 F1$ = F$(1,S- 1) \ REM FIRST NAME
400 F2S = F$(S + 1 ,C - 1) \ REM LAST NAME
410 F$ = F2$+“,” + Fl$
415 REM PUT BACK IN MATRIX (NOTE FULL 30CHARS SO NO LEFT-OVERS)
420 N$(FNL(N1), FNH(N1)) = FNA$(F$)
430 NEXT N1
500 REM BUBBLE SORT, LOOP UNTIL NO SWAP ON A PASS
510 F = 0
520 FOR 1 = 2 TO N9
530 IF N$(FNL(I),FNH(I))> = N$(FNL(I - 1),FNH(I - 1)) THEN 590
540 REM SWAP
550 F = 1 / REM REMEMBER A SWAP WAS DONE
560 F$ = N$(FNL(I),FNH(I))
570 N$(FNL(I),FNH(I)) = N$(FNL(I- 1),FNH(I- 1))
580 N$(FNL(I - 1),FNH(I — 1)) = F$
590 NEXT I
600 IF FX) THEN 510 \ REM KEEP TRYING TILL NO SWAPS
800 REM PRINT SORTED LIST
805 PRINT \ PRINT \ PRINT “ ♦*** SORTED NAMES ***♦” \ PRINT \ PRINT
810 FOR 1 = 1 TO N9
820 F$ = N$(FNL(I),FNH(I))
830 PRINT F$
840 NEXT I
850 END
READY
Program A. Mailing list (HP style).
97
MAILING. BAS WRITTEN IN BASIC-E (11/6/77)
5 REM WRITTEN IN BASIC-E
7 REM GET NUMBER OF NAMES
10 READ N9
15 DIM N$(N9)
200 REM READ IN NAMES
205 PRINT “ **** NAMES ♦***”
210 FOR 1 = 1 TO N9
220 READ N$(I)
225 PRINT N$(I)
230 NEXT I
240 DATA 10
250 DATA SALLY JONES, SAM SMITH, JOE SMITH, TIM CAMBELL, ED HILL
260 DATA STEVE MOODY, ROGER HEAD, SHIRLEY JONES, ISSAC DEAR, RICH KING
300 REM RE-ORDER LAST NAME FIRST
310 FOR N1 = 1 TO N9
320 C = 1 : F$ = N$(N1) : L = LEN(F$)
325 REM LOOP UNTIL LAST CHAR AND MARK LAST BLANK
330 IF OL THEN 365
340 IF MID$(F$,C1) = “b” THEN S = C
350 C = C+1
360 GOTO 330
365 REM ACTUALLY SHUFFLE NAMES
370 Fl$ = LEFT$(F$,S - 1) : REM FIRST NAME
379 REM NOTE RIGHT$(NAME, LENGTH)
380 F2$ = RIGHT$(F$,L - S) : REM LAST NAME
390 F$ = F2$ + ‘\” + Fl$
392 REM FILL OUT LENGTH SINCE 3 CHAR STR<4 CHAR STR
395 N$(N1) = F$ + LEFT$(“ ”,30-LEN(F$))
400 NEXT N1
500 REM DO SIMPLE BUBBLE SORT
510 F = 0 : REM LOOP UNTIL NO SWAPS ON A PASS
520 FOR 1 = 2 TO N9
530 IF N$(I)>N$(I - 1) THEN 590
540 REM SWAP
550 F= 1 : REM REMEMBER SWAP
560 F$ = N$(I)
570 N$(I) = N$(I — 1)
580 N$(I — 1) = F$
590 NEXT I
600 IF FX) THEN GOTO 500 : REM TEST FOR DONE
800 REM PRINT SORTED LIST
810 PRINT : PRINT : PRINT “ *•** SORTED LIST
820 FOR I = 1 TO N9
830 PRINT N$(I)
840 NEXT I
Program B. Mailing list (DEC style).
ferent ways of using strings in
BASIC, both are common
enough to have a following, but
the most useful one is the one
on your computer. Which is bet-
ter? It’s not for me to know;
however, I have used both long
enough to know that strings
make a program really fun to
use— even if it’s a business pro-
gram. That is because we talk
in strings, not numbers. Like
other computer users, I have
braved strings in FORTRAN
(which has no strings) and
thrilled to a real string lan-
guage like SNOBOL (running
on a 360/65 in 250K). You use
what you have! And hope some-
one’s coming along with some-
thing better. Until then, keep on
coding!!
References
1. Data General Extended
BASIC User’s Manual, Rev 6,
Feb. 1975.
2. DEC PDP-11 BASIC- PLUS
Language Manual, July 1975.
3. Altair BASIC Reference Man-
ual, 1975.
4. Tymshare BASIC Tycom-X
Manual, March 1973.
5. CTSTS BASIC Reference (IN-
FONET, Computer Sciences
Corp.), May 1974.
6. Timeshare BASIC/2000 Level
F Reference Manual (Hewlett-
Packard), Feb. 1975.
7. North Star BASIC Version 6
Manual, Feb. 1977.
8. Personal Notes from GE-635
Mark I Timesharing, Oct. 1968.
**** NAMES **♦*
ters of a name element of 30
characters in the pseudomatrix
N$ of names. FNA$ is used to
fill a name out to 30 characters.
Since a pseudomatrix element
must be a fixed length, $ is used
at the end of a name on initial
entry so the first name/last
name swap tells where the
name ends.
The DEC-style version looks
much nicer, primarily because
it accepts a tab character while
being typed in and thus is
easier to format (called pretty-
print). It is wise to do this if you
can since it makes the reading
of the program easier.
Line 380 uses the RIGHTS
the length, so RIGHTS returns
the right n-most characters
(i.e., RIGHT$(“ABCDEF”,3) =
“DEF”). Yet a true DEC-written
BASIC will return from the nth
character to the end (i.e.,
RIGHT$(“ABCDEF”,3) =
“CDEF”). Line 395 illustrates
one of the nice things about a
DEC-like BASIC— string ele-
ments of variable length. This
particular BASIC says a long
string of As is greater than a
short string of Bs, i.e., AAA>
BB. Well, to each his own.
(Note: This is specific to this
BASIC (BASIC-E), not to all
DEC-like BASICS.)
SALLY JONES
SAM SMITH
JOE SMITH
TIM CAMBELL
ED HILL
STEVE MOODY
ROGER HEAD
SHIRLEY JONES
ISSAC DEAR
RICH KING
**♦* SORTED NAMES ****
CAMBELL, TIM
DEAR, ISSAC
HEAD, ROGER
HILL, ED
JONES, SALLY
JONES, SHIRLEY
KING, RICH
MOODY, STEVE
SMITH, JOE
SMITH, SAM
READY
function. This particular BASIC Summary
Fig. 1. List sort.
has the second parameter as We have looked at two dif-
98
• •
YOUR BEST BUYS ARE AT
8v@30A, ±16v@10A option
SS-50 bus option
voltage monitor option
aniputer
ft ruer Jersey I
240 WANAOUE AVE.,
P0MPT0N LAKES, N.J. 07442
(201) 835-7080
Telex-130- 376 amcaTl pomt
micro-mini computer center
STORE HOURS: MON.-FRI. 6:30p.m. tolOp.m. SAT.10 a.m. to 6 p.m.
AUTHORIZED DEALER FOR:
• Southwest Technical Products Corp. • Seals Electronics, Inc.
•Jim-Pak Electronic Components • PolyMorphic Systems •Micro-Term
• Ohio Scientific • Technico, Inc. • Hayden Books • Kim-1 • Sanyo
LOOK AT THESE PRICES!
Only at Computer Corner of N.J.
TECHNICO TMS 9900-16 Bit Processor
Super Starter System T-9900-SS-U
Regularly $299 NOW $ 269® 5
Also the
New Netronics ELF II, RCA COSMAC microprocessor,
minicomputer for only $139.95 wired & tested!
VERBATIM Removable Magnetic Storage Media
Minidiskettes md52«>i MD525-10 ' 9 1025 26-100
MD 525-16 *-50 4 25 3.75
Standard Size FD34-1000 FD32-1000 19 1025 26-100
Diskettes FD65-1000 s.so 5.2s 475
TO ORDER:
Send check or money order and include $1.00 for shipping, 1.00 (optional) for
insurance, and please include 5% sales tax if you are a New Jersey state
resident. Phone orders accepted with Mastercharge or Visa.
kack
mounted
model $200
Desk
top model
$235
Write or call for a copy of our
detailed brochure which includes
our application note
BUILDING CHEAP COMPUTERS.
INrEGR4ND
8474 Ave. 296 • Visalia. CA 93277 • (209) 733-9288
We accept BankAmericard/Visa and Master Charge
NORTH STAR
DISK ASSEMBLER
and
DISK EDITOR
Both programs read and write disk files; file size not
limited by memory. Assembler will assemble up to
ten source files at a time; permits modular pro-
gramming with programs easily relocated by
reassembling at the desired address. Editor does
not use line numbers; it searches for strings. Lines
may be inserted, deleted and displayed. Large disk
source files allow programs to be fully commented.
ASSEMBLER/EDITOR on disk
with users manual. ..$30
COMPUTER SYSTEMS DESIGN
1611 E. Central Wichita, KS 67214
DEALERS INQUIRIES INVITED
C46 J
16K Static RAM
S-100 Systems
NEW
LOW PRICES
*330 Kit
*365 Assembled
• Very Low Power-650MA+5V; 90MA+12V;
16MA-5V
• Applications Notes— 6800 and 6502 Sys.
• Low-profile sockets for all chips
• Solder mask; silk screen; plated through holes
• Each 4K addressable to any 4K boundary
• Fully buffered S-100 bus— gold-plated contacts
• NEC juPD 410 D memories
COD, Master Charge, B of A, Visa Accepted ....
Orders shipped prepaid. California residents
add 6% sales tax.
ANDENBERG DATA PRODUCTS
PO BOX 2507
SANTA MARIA CALIFORNIA 93454
•05-937-7951
V14
99
Thomas E. Doyle
5222 Big Bow Road
Madison Wl 53711
5 Minutes
or 5 Hours?
sorting techniques compared
Time Required to Sort N Items (seconds)
N
Ripple
Modified
Bubble
S-M
50
61
43
33
9
100
245
173
130
21
150
552
390
290
36
300
-
-
1224
85
Number of Swaps of
Entries
N
Ripple
Modified
Bubble
S-M
50
1225
1225
1225
105
100
4950
4950
4950
260
150
11175
11175
11175
425
300
-
-
44850
1000
Number of
Entry Comparisons
N
Ripple
Modified
Bubble
S-M
50
2450
1225
1225
263
100
9900
4950
4950
668
150
22350
11175
11175
1187
300
-
-
44850
2812
Table 1.
I n an attempt to help justify
the purchase of a floppy-disk
system, I decided to put the
computer to some practical
use. It seems that not everyone
considers piloting the Enter-
prise and destroying Klingons
as a useful function worthy of
another kilobuck investment.
Using the system to keep track
of household expenses seemed
to be a good place to start. The
Do-All program by Randy Miller
( Kilobaud , August 1977) provid-
ed an ideal program.
After the program was load-
ed, a list of about a hundred
items was entered for my
demonstration of the practical
advantages of a home com-
puter. Everyone gathered for
the show, and the program was
run. A command was given to
sort the list of data alphabeti-
cally. Everyone stared at the
printer waiting for the output
from this electronic marvel.
Nothing happened.
Taking advantage of the
pause and the presence of a
captive audience, I discussed
the advantages of adding a
disk to the wonderful com-
puter. At the end of my rather
lengthy discussion there was
still nothing on the printer. As
time wore on, I began to con-
sider the possibilities: hard-
ware problems, software prob-
lems or simply another exam-
ple of Murphy’s Law. I felt there
must be something wrong.
After all, the Enterprise could
move across the entire galaxy
in only seconds, so alphabetiz-
ing this list could not take that
long. Trying to remain cool, I
suggested that we leave the
computer and come back when
it was done.
Much to my suprise, thirty
minutes later the sorting was
100
5 REM --- RIPPLE SORT ---
6 REM — SET UP ARRAY
10 N = 1 50
20 DIM DC N)
30 J = N
40 FOR 1=1 TO N
50 DC I ) =J
60 J=J-1
70 NEXT
80 PRINT
90 REM --- START OF SORT ---
100 M=N
10 5 C=0
110 FOR 1=1 TO M-l
120 CM=C M+l
130 IF DCI) <=DCI+1 ) THEN 1 60
135 SW=SW+1
140 T= DC I ) : DCI) =DCI+1 )s DCI+l ) =T
150 C=1
160 NEXT I
170 IF C=1 THEN 105
300 REM --- PRINT RESULTS ---
31 0 PRINT "SWI TCHES =" ;SW
320 PRINT "COMPARISONS =” ;C M
330 PRINT "SIZE -" ;N
OK
Program A.
complete. The printout re-
vealed that the list had been
sorted exactly as requested.
What could have caused the
delay? Perhaps my 8080 was
slow. The benchmark programs
in the basic timing com-
parisons article ( Kilobaud ,
June 1977) were run and re-
vealed that my computer ran a
little faster than the one used
for the article.
Since the program ran prop-
erly and the computer was up
to speed, the solution to the
problem must be in the sorting
technique used in the program.
An article on sorting routines
by Andrew J. Rerko (Kilobaud,
April 1977) was consulted and
some test programs (Programs
A, B and C) were run using the
Ripple, Modified Ripple and
Bubble routines described in
the article.
The test programs consisted
of setting up an array of N
numbers in reverse order and
using each of the sorting
routines to sort them. The pro-
gram execution times as well
as number of comparisons and
the number of element switch-
es were recorded. The results
are shown in Table 1. The
results of this test revealed two
things: The bubble sort was a
little faster than the others, and
sorting takes a lot of time. Sort-
ing a simple table of 100
numbers took almost three
minutes. No wonder the Do-All
program took so long.
None of the common sorting
methods described in Mr.
Rerko’s article would speed up
a sorting program significantly.
The solution to the problem, if
any, would lie in an uncommon
sorting routine. An article by
John P. Grillo (Creative Com-
puting, November 1976) dis-
cusses a technique called the
Shell-Metzner Sort. This
method offered significant
speed advantages when sort-
ing large amounts of data. A
flowchart of the Shell-Metzner
Sort is shown in Fig. 1 . The arti-
cle stated that a projected sort
of 10,000,000 items would take
93 years using a bubble sort.
Using the S-M technique, sort-
ing the same data would re-
quire only 2.5 days. But would it
help when sorting small
amounts of data?
The benchmark sorting pro-
5 REM --- MODIFIED RIPPLE SORT
6 REM — SET UP ARRAY —
10 N = 1 50
20 DIM D(N)
30 J=N
40 FOR 1=1 TO N
50 DC I ) =J
60 J=J-1
70 NEXT
80 PRINT
90 REM --- START OF SORT ---
100 M=N
110 C=0
112 M=M-1
11 5 IF M=0 THEN 300
120 FOR 1=1 TO M
125 C M=C Wf 1
130 IF DCI) <=D( 1+1 ) THEN 1 60
135 SW=SW+1
140 T=DCI):DCI)=DCI + l ): DCI+l) =T
150 C=!
160 NEXT I
170 IF C=l THEN 1 10
300 REM PRINT RESULTS ---
310 PRI NT "SWITCHES =" ;SW
320 PRINT "COMPARISONS =" ;C M
330 PRINT "SIZE -" ;N
OK
Program B.
5 REM --- BUBBLE SORT ---
6 REM --- SET UP ARRAY ---
10 N=1 50
20 DIM DC N)
30 J = N
40 FOR 1=1 TO N
50 DC I ) =J
60 J=J-1
70 NEXT
80 PRINT "*"
90 REM --- START OF SORT ---
100 M=N
110 FOR 1 = 1 TO M-l
120 FOR J=I + 1 TO M
125 CM=CWfl
130 IF DCI) < = DC J ) THEN 170
135 SW=SW+1
140 T= DC I ) : DCI) = DCJ): DCJ) = T
170 NEXT J
180 NEXT I
300 REM PRINT RESULTS ---
310 PRI NT "SWI TCHES =" ;SW
320 PRI NT " COMPARISONS =" ;C M
330 PRI NT "SIZE -" ; N
OK
Program C.
gram was run using the S-M
method and is shown in Pro-
gram D. When sorting 150
items, the S-M sort was over
eight times faster than the bub-
ble sort and over 15 times faster
than a ripple sort. The bubble
sort required over 20 minutes to
101
Fig. 1. Shell-Metzner Sort.
sort 300 items. The S-M method
required only 85 seconds to
sort the same list. The speed
advantage of the S-M sort in-
creases dramatically with the
size of the list, but it seemed to
speed sorts of even small lists.
The next step was to incor-
porate the S-M sort technique
into the Do-All program and try
it out. A random list of 100 en-
tries was prepared and sorted
by the standard program. Al-
most 45 minutes were required
to sort this list. The Do-All pro-
gram was then modified to use
the S-M sort. Sorting the same
list of 100 entries now required
less than nine minutes. To
modify the Do-All program,
remove lines 4050-4115,
4150-4280, 9220-9340 and
replace with the new lines
shown in Progam E.
The only disadvantage I have
found with the S-M technique
so far is that it does require
slightly more code, and it uses
five index variables rather than
5 REM SHELL METZNER SORT
6 REM --- SET UP ARRAY —
10 N=300
20 DIM D(N)
30 J=N
40 FOR 1=1 TO N
50 D( I ) =J
60 J=J-1
70 NEXT
80 PRINT
90 REM START OF SORT
100 M=N
110 M = I N T( M/2 )
120 IF M=0 THEN 300
130 J=1 : K =N-M
1 40 I =J
150 L= 1 + M
1 55 CM=CMf 1
160 IF DC I) <D( L) THEN 210
170 T=D ( I ) : D(I) = D(L): D(L) = T
175 SW=SW+1
180 I = I -M
190 IF I<1 THEN 210
200 GOTO l 50
210 J=J+1
220 IF J >K THEN 110
230 GOTO 1 40
300 REM --- PRINT RESULTS ---
310 PRINT "SWI TCHES =" :SW
320 PRINT "COMPARISONS =" ;C M
330 PRINT "SIZE -" ;N
OK
Program D.
only one or two as other sorting
methods. Following the exam-
ple benchmark program, it
should be possible to use the
S-M technique in other sorting
programs.
Notes on Programs
All programs were run on an
8080 system with a 2 MHz clock
and zero wait states. Mits 8K
BASIC (Version 3.2) was used.
Variable CM was used to total
the number of comparisons be-
tween table entries. The vari-
able SW was used to total the
number of switches between
table entries.!
LIST
4050
4050
M = P
4055
M=I NT( M/2)
4060
IF M=0 THEN 1 1 40
4065
J=1 J K=(P-1)-M
4070
I=J
407 5
L=I+M
4080
IF N( T,I) < = N( T,L)
THEN 4105
408 5
G0SUB 9210
4090
I=I-M
409 5
IF I<1 THEN 4105
4100
GOTO 407 5
4105
J=J+1
4110
IF J >K THEN 40 55
41 1 5
GOTO 4070
BREAK
OK
LIST
41 50
41 50
M=P
41 60
l*!=I NT( IV2)
4170
IF M=0 THEN 1 I 40
4180
J = 1 S K=(P-1)-M
4190
I =J
4200
L=I+M
421 0
IF A$( T,I) <=A$( T,
L) THEN 42 60
422 0
G0SUB 9210
42 3 0
I =I-M
42 40
IF I<1 THEN 42 60
42 50
GOTO 4200
42 60
J -J+I
427 0
IF J>K THEN 4160
4280
GOTO 4190
BREAK
0K
LIST
9220
9220
XI =N< 1 , L)
9230
X2=N(2,L)
92 40
B 1 S=AS( I , L)
92 50
B2$=A$<2 , L)
92 60
FOR Z=1 TO 2
92 7 0
N(Z,L) =N(Z,I>
9280
A$(Z,L) =A$(Z,I)
92 9 0
NEXT
9300
N(1 ,1) =X1
9310
N(2,I) =X2
9320
A$( 1 , 1 ) =B 1 $
9330
A$(2 ,1) =B2$
93 40
RETURN
BREAK
OK
Program E.
102
® o o :
AMPLIFIER *
\J
S-100 BUS
t
CSR 1
SYNTHESIS-BY-RULE
SOFTWARE
t
SPEAK “KAAMPYUTAOLKER”
MODEL CT-1 SYNTHESIZER 395.00
CSR1 SOFTWARE SYSTEM 35.00
DEMONSTRATION CASSETTE 2.95
CALIF. RESIDENTS ADD 6% SALES TAX
WRITE FOR INFORMATIVE LITERATURE
COMPUTALKER CONSULTANTS
P.O. BOX 1951, DEPT. K, SANTA MONICA, CA 90406
C36
MINATURE SOLID STATE
202 VIDEO CAMERA KIT
FEATURING A ... 100 x 100 BIT SELF SCANNING CHARGED COUPLED DEVICE
THIS UNIQUE UPDATED CAMERA KIT
FEATURES THE FAIRCHILD CCD 202C IMAGE SENSOR
FEATURES
ADVANTAGES
We supply all semiconductors, boards, data sheets, $75 00 to assemble
diagrams, resistors, capacitors, and 8MM lens. and test— add $2.00
Sorry we do not supply the case, batteries and 5V supply. posta 9 e & handling.
S5
• IN THE FUTURE
WE WILL SUPPLY A
COMPUTER VIDEO INTERFACE CARD
• All clock voltages operate at 6V
reguiring no adjustments
• Higher video output signal
• We supply the power board, so only
a 5V 1 Amp power source is needed
• The circuitry has been simplified for
easier assembly
• Two level TTL output is supplied for
interfacing
*349°°
Sensitive to infra red
as well as visible light
May be used in a vacuum,
under water and high altitude
May be used in a magnetic
environment because there is
no high voltage
All components mounted on
parallel 3 3 /4 ff x 6V2" single
sided boards
Total weight under 1 lb.
0
SOLID STATE
P0B0X74K SOMERVILLE, MASS. 02143
SALES
TEL (617)547-7053
Micro-Diagnostic® Extensive documentation, Fully Socketed.
Piebug Monitor: Relative address calculator. Pointer High-low,
User Subroutines, Back-step key.
Cassette Interface: Load S Dump by file *, Tape motion
control , Positive indication of operation .
Applications systems from $90 (lOunit quantity)
Development systems from $149 (single unit)
TELL ME MORE ;
( ) Please send documentation
and price lists. $10 enclosed.
( ) I don' t need documentation
please send price lists.
{ ) Please send FREE CATALOG.
want to see for myself that the 8700 is The Answer.
address:
Citv:
state:
ziD:
Wilshire Blvd. •
Oklahoma City. OK 73116
(405) 843-9626
KIMSI 1
The KIM to S-100 bus
Interface/Motherboard
• Combines the power of the 6502 with the flexibility of
the S-100 bus
• Attaches to any unmodified KIM
• Complete interface logic and fully buffered motherboard
in one unit
• On-board regulation of power for KIM
• Eight slots of S-100 compatibility for additional RAM,
Video and I/O boards, PROM Programmers, Speech
processors . . .
• Includes all parts, sockets for ICs, one 100 pin connector,
and full Assembly/Operating documentation
♦ Kit $125, Assembled $165
♦ All units shipped from stock
FORETHOUGHT PRODUCTS
P.O. Box 8066-H mm
Coburg, OR 97401
ZJ
103
Do-It-Yourself Time-sharing
it’s easier than you think
Mike Kop
3060 Marshall Ave.
Cincinnati OH 45220
W hen I first learned to pro-
gram I was taught how to
sign on to a computer system
using a teletypewriter, type in a
program and obtain the results
at the terminal almost im-
mediately. Other users around
me, each working on his own
program, were using similar
type terminals. It appeared that
each user had the entire com-
puter to himself! This amazed
and perplexed me. How could a
computer run all the terminals
and keep track of every-
body— all at the same time? I
conceded that the system was
too complex to analyze (or
perhaps it was sheer magic).
Eventually I began to under-
stand what went on by fighting
my way through books on
operating systems. I hope that
future computer users will be
spared a similar experience.
Last year, I purchased an
M6800 system from SWTP.
After programming on it for a
while, I decided to investigate
the possibility of implementing
time-sharing on my system. It
turns out to be simpler than you
might think.
In this article, I will attempt
to explain exactly why one
would want to set up time-
sharing and how it is done (for
an M6800 system). I’ll also try to
explain some other program-
ming considerations.
What is Time sharing?
Time-sharing is accom-
plished by switching rapidly be-
tween many users. That means
each user is allowed, in turn, a
short duration of central pro-
cessing unit (CPU) or micropro-
cessing unit (MPU) time. This is
called a time slice. For exam-
ple, if the time slice were 50
milliseconds, then each user
would use the processor for 50
milliseconds. If the switching is
fast enough, the computer op-
eration from each user’s point
of view will appear continuous.
Why Time-sharing?
The computer in a large
system may cost several
million dollars. Obviously, buy-
ing one computer for each user
is extremely impractical. Shar-
ing the computer among many
users is a more effective way to
utilize the system.
Another reason for time-
sharing is because a com-
puter’s input/output (I/O)
devices are much slower than
the processor. If a terminal is
outputting characters at 30
cps, there is sufficient time be-
tween characters for other
work. Thus, with time-sharing,
literally two, three or more
times as much work can be ac-
complished than by a single
user.
Most of the reasons given for
using time-sharing would also
apply to a microcomputer
system (perhaps on a smaller
scale). One possible argument
against its use in microproces-
sors would be that they’re too
slow. However, for programs
that do a lot of input and output
and use little processor time
(most games and business-
type programs fall into this
category), I see no reason why
time-sharing cannot be im-
plemented.
Using Interrupts
Proper use of interrupts
comes first in implementing
time-sharing. The ideas
presented here are essentially
the same, whether you have a
small or large system.
An interrupt is basically a
hardware mechanism that
makes the microprocessor
stop what it is doing and jump
to another program (often
known as a service routine).
Sometimes it is possible to
mask off an interrupt. If this
happens, then the interrupt is
ignored (or held pending until
some later time).
Let’s look briefly at the inter-
rupt mechanisms on the SWTP
system (which uses MIKBUG).
There is a line marked IRQ (for
interrupt request). If this line is
temporarily gounded and the
mask bit is a zero, an interrupt
will occur. The system will then
jump to the address contained
in storage locations $A000 and
$A001 . One nice thing about the
M6800 microprocessor is that
when interrupted it stores
everything (i.e., the condition
code, B, A, X and program
counter registers) on the stack.
This means that little effort is
required to remember where
each program is when it was
stopped. With other proces-
sors, you would typically have
to store all registers away,
which may take many instruc-
tions. One danger of this is that
if another interrupt occurs
before all registers are stored
away, some register contents
may be lost. The M6800 proces-
sor saves everything in one
swoop.
104
Incidentally, you may, if
desired, use the nonmaskable
interrupt NMI instead of IRQ.
The interrupt address would
then be stored at locations
$A006 and $A007. I prefer,
however, to use an interrupt
that is maskable.
Software
Program A actually imple-
ments time-sharing. The com-
ments should aid you in under-
standing how the program
works. It starts at address
BEGIN. Also, some hardware
must be set up so that an IRQ
interrupt is generated at
regular intervals (this is ex-
plained later). Each time an in-
terrupt is generated, one pro-
gram is stopped and the next
one in line is started. For exam-
ple, if program 1 is currently ex-
ecuting and we are time-
sharing three programs, then
four interrupts will result in pro-
gram 2 being executed (1 then 2
then 3 then 1 then 2). With Pro-
gram A, you may time-share up
to 15 different programs.
The part of the program that
actually does the time-sharing
(the service routine) is
statements 69 to 83. State-
ments 1 to 64 merely initialize
various parameters. The in-
itialization routine basically
works thus— initially each pro-
gram is assigned a stack point-
er. The stack-pointer address-
es differ by 16 bytes. That is,
program 1 has a stack-pointer
value of END + 16, program 2
has a value of END + 32, etc.
These values are stored at ad-
dresses STACK1, STACK2, etc.
The initialization routine also
clears the condition-code regis-
ter and stores the starting ad-
dress of each program at the
appropriate position in each
stack. When the RTI instruction
is executed, the processor
fetches all registers (program
counter included in the fetch)
from the stack and starts (or
resumes) a program at the ap-
propriate address.
The purpose of clearing the
condition code in the stack for
each program is that when the
RTI instruction is executed, the
interrupt mask bit will not
become set (which would lock
up the system). For example, if
the stack pointer were at $0F00,
we would clear address $0F01
and store the starting address
at address $0F06. An RTI in-
struction would then load the
condition-code register with
$00 and the program counter
with the number at address
$0F06. Initially, we don’t care
what the other register con-
tents are.
The service routine performs
a very simple function. It stops
the current program from ex-
ecuting and runs the next pro-
gram in line; it accomplishes
this by storing away the current
stack pointer and loading the
next one. When the RTI instruc-
tion is executed, we do not
return exactly where we left off
(that is, resume execution of
the same program) as is nor-
mally done. Instead, we go to
the next program. This occurs
because the stack pointer has
been changed.
You will also observe that in
the service routine, I purposely
store data where instructions
are. This is a trick I use to make
the service routine execute
quickly, although in general
this is not good practice. I do
have another version of the ser-
vice routine that does not do
this; however, it is slightly
longer.
For a simple demonstration
of time-sharing, Program B may
be used. This program
assumes that you have a serial
interface port (which uses an
ACIA) at the correct baud rate
at address $8008. You will also
have to have a terminal plugged
in at this address. We will call
this terminal 2. Terminal 1 will
be at the control interface. If
you run the Program B starting
Program A. Time-share program.
STMT
ADDR
CODE
STATEMENT
1
0E00
ORG
$0E00
2
0E00
STRING
EQU
$E07E
3
0E00
IN2HEX
EQU
SE055
4
0E00
INHEX
EQU
$E0AA
5
0E00
CR
EQU
SOD
6
0E00
LF
EQU
$0A
7
0E00
EOT
EQU
$04
8
0E00
CE
0E92
BEGIN
LDX
#SERVCE
9
0E03
FF
A0O0
STX
SA000
INITIALIZE INTERRUPT REQUEST POINTER
10
0E06
CE
0E61
LDX
#MES1
11
0E09
BD
E07E
JSR
STRING
PRINT ‘#PROGRAMS = ’
12
0E0C
BD
E0AA
JSR
INHEX
GET NUMBER OF PROGRAMS TO BE TIME SHARED
13
0E0F
B7
0EB2
STA
A
NUMBER
14
0E12
16
TAB
15
0E13
CE
0EB3
LDX
#STACK1
16
0E16
FF
0E5D
STX
TEMPO
17
0E19
CE
0EE1
LDX
#END+ 16
X-REG NOW POINTS TO THE BEGINNING
18
0E1C
FF
0E5F
STX
TEMP
OF THE STACK AREA
19
0E1F
FE
0E5D
A1
LDX
TEMPO
LOAD ADDRESS OF STACK I
20
0E22
FF
0E2E
STX
ST0+1
21
0E25
08
INX
22
0E26
08
INX
23
0E27
FF
0E5D
STX
TEMPO
STORE ADDRESS OF STACK I + 1
24
0E2A
FE
0E5F
LDX
TEMP
25
0E2D
FF
FFFF
ST0
STX
SFFFF
INITIALIZE STACK I
26
0E30
6F
01
CLR
1,X
CLEAR CONDITION CODE REGISTER I
27
0E32
86
06
LDA
A
#6
28
0E34
8D
22
BSR
ADD
29
0E36
FF
0E4A
STX
ST + 1
THE X-REG NOW POINTS TO THE ADDRESS WHERE THE
30
0E39
86
0A
LDA
A
#10
STARTING ADDRESS OF PROGRAM I STARTS
31
0E3B
8D
IB
BSR
ADD
32
0E3D
FF
0E5F
STX
TEMP
THE ADDRESS OF THE NEXT STACK WILL BE 16
33
0E40
CE
0E73
LDX
#MES2
BYTES AWAY FROM THE CURRENT STACK
34
0E43
BD
E07E
JSR
STRING
PRING ‘START = *
35
0E46
BD
0E7C
JSR
INPUTX
INPUT STARTING ADDRESS
36
0E49
FF
FFFF
ST
STX
SFFFF
INITIALIZE PROGRAM COUNTER 1
37
0E4C
5A
DEC
B
38
0E4D
26
DO
BNE
A1
105
39
0E4F
86
01
LDA
A
#1
BEGIN RUNNING PROGRAMS
40
0E51
B7
0EB1
STA
A
STATUS
41
0E54
BE
0EB3
LDS
STACK 1
42
0E57
3B
RTI
43
OE58
08
ADD
INX
THIS SUBROUTINE INCREASES THE X-REG
44
0E59
4A
DEC
A
BY THE VALUE IN THE A-REG
45
0E5A
26
FC
BNE
ADD
46
0E5C
39
RTS
47
0E5D
TEMPO
RMB
2
48
0E5F
TEMP
RMB
2
49
0E61
ODOA
MES1
FCB
CR,LF
50
0E63
2350
FCC
/#PROGRAMS(l-F)?/
524F
4752
414D
5328
4
312D
4629
3F
51
0E72
04
FCB
EOT
52
0E73
ODOA
MES2
FCB
CR,LF
53
0E75
5354
FCC
/START = /
4152
543D
54
0E7B
04
FCB
EOT
55
0E7C
36
INPUTX
PSH
A
SUBROUTINE TO INPUT THE X-REG
56
0E7D
37
PSH
B
57
0E7E
BD
E055
JSR
IN2HEX
58
0E81
B7
0E90
STA
A
DATA
59
0E84
BD
E055
JSR
IN2HEX
60
0E87
B7
0E91
STA
A
DATA + 1
61
0E8A
FE
0E90
LDX
DATA
62
0E8D
33
PUL
B
63
0E8E
32
PUL
A
64
0E8F
39
RTS
65
0E9Q
DATA
RMB
2
66
*
67
♦THE PURPOSE OF THIS SERVICE ROUTINE IS TO
68
♦STOP PROGRAM I AND BEGIN RUNNING PROGRAM I + 1
69
0E92
CE
0EB1
SERVICE
LDX
#STACKl-2
70
0E95
B6
0EB1
LDA
A
STATUS
DETERMINE THE PROGRAM CURRENTLY EXECUTING
71
0E98
48
ASL
A
72
0E99
B7
0E9D
STA
A
ST1 + 1
73
0E9C
AF
00
ST1
STS
X
SAVE THE STACK POINTER AT THE APPROPRIATE
74
0E9E
47
ASR
A
ADDRESS(ST ACK 1 ,STACK2, . . ..STACKF)
75
0E9F
4C
INC
A
BEGIN TO EXECUTE THE NEXT PROGRAM
76
0EA0
B1
0EB2
CMP
A
NUMBER
CHECK FOR WRAP AROUND
77
0EA3
2F
02
BLE
L3
78
0EA5
86
01
LDA
A
#1
IF WRAP AROUND EXISTS EXECUTE PROGRAM #1
79
0EA7
B7
0EB1
L3
STA
A
STATUS
INDICATE THAT THE NEXT PROGRAM IS EXECUTING
80
OEAA
48
ASL
A
81
OEAB
B7
OEAF
STA
A
ST2+1
82
OEAE
AE
00
ST2
LDS
X
LOAD THE APPROPRIATE STACK POINTER
83
OEBO
3B
RTI
BEGIN ACTUAL EXECUTION
84
0EB1
STATUS
RMB
1
CURRENT PROGRAM IN EXECUTION 1 TO F)
85
0EB2
NUMBER
RMB
1
TOTAL NUMBER OR PROGRAMS TO BE TIME SHARED
86
0EB3
STACK 1
RMB
2
STACK POINTER FOR PROGRAMS
87
0EB5
STACK2
RMB
2
STACK POINTER FOR PROG RAM #2
88
0EB7
STACK3
RMB
2
. . . ETC . . .
89
0EB9
STACK4
RMB
2
90
OEBB
STACK5
RMB
2
91
OEBD
STACK6
RMB
2
92
OEBF
STACK7
RMB
2
93
0EC1
STACK8
RMB
2
94
0EC3
STACK9
RMB
2
95
0EC5
STACKA
RMB
2
96
0EC7
STACKB
RMB
2
97
0EC9
STACKC
RMB
2
98
OECB
STACKD
RMB
2
99
OECD
STACKE
RMB
2
100
OECF
STACKF
RMB
2
101
*
102
0ED1
END
EQU
*
103
*
104
A048
ORG
$A048
105
A048
OEOO
FDB
BEGIN
SYMBOL
VALUE
DEFN
REFERENCES
STRING
E07E
2
11 34
106
at address $0000, a series of
zeros should be printed out on
terminal 2. Starting at address
$0008 will result in a printout of
ail ones.
We will now time-share both
parts of this program. For this
part, first press the reset but-
ton. This will set the mask bit to
a one. Now set the interrupt
rate to a very slow value, say
once every ten seconds if
possible. (We’ll discuss the
hardware to accomplish this in
a moment.) Now run Program A,
starting at address BEGIN
($0E00). You will then be re-
quired to type in the number of
programs you want (this is a
single hex number from 1 to F)
to time-share, followed by their
respective starting addresses.
The data is entered as follows:
#PROGRAMS(l-F)?2
START = 0000
START = 0008
After having done the above,
you should see the printout at
terminal 2 alternate between
strings of zeros and strings of
ones. If you slowly increase the
interrupt rate you will notice
that the respective strings
become shorter and shorter.
If you do not have a second
terminal, you may unplug the
terminal from the control inter-
face in each of the above steps
and plug it into the other port
after having typed a G. Be very
careful when doing this; you
should avoid the practice in
general.
Perhaps you have wondered
why I used another I/O port and
not MIKBUG directly. MIKBUG
outputs a character by soft-
ware, bit by bit. If you were to in-
terrupt the output routine, the
output bits would not appear at
the proper time. That is, you
cannot output part of a charac-
ter now and the other part later.
This problem does not occur
with an ACIA because a
character is output by a single
store instruction.
Hardware
As stated previously, inter-
rupts must be generated at
regular intervals. An interrupt
should be generated by a pulse
that grounds the IRQ line for a
very short duration before
returning to a high state. This is
because the IRQ line must
return to its high state before
the service routine has com-
pleted its job. If this is not done,
then another interrupt will oc-
cur immediately after the ser-
vice is completed, causing
some programs to be skipped
in execution. A pulse duration
of 50 microseconds works quite
well. An interrupt will not occur
inside the service routine
because the mask bit will be set
at that time. If, however, you
decide to use NMI instead, your
pulse must be much narrower
(e.g., 10 microseconds). Other-
wise, the service routine may
keep interrupting itself, which
can lead to difficulties!
If you have a signal generator
that can generate a pulse, so
much the better. I also under-
stand that SWTP now has
available an interrupt timer
board. In place of these alter-
natives, you may use the circuit
shown in Fig. 1. There are no
doubt other circuits that will
work as well. Resistors R1 and
IN2HEX
E055
3
57
59
INHEX
E0AA
4
12
CR
000D
5
49
52
LF
000A
6
49
52
EOT
0004
7
51
54
BEGIN
0E00
8
105
A1
0E1F
19
38
ST0
0E2D
25
20
ST
0E49
36
29
ADD
OE58
43
28
31
45
TEMPO
0E5D
47
16
19
23
TEMP
0E5F
48
18
24
32
MES1
0E61
49
10
MES2
0E73
52
33
INPUTX
0E7C
55
35
DATA
0E90
65
58
60
61
SERVCE
0E92
69
8
ST1
0E9C
73
72
L3
0EA7
79
77
ST2
0EAE
82
81
STATUS
0EB1
84
40
70
79
NUMBER
0EB2
85
13
76
STACK 1
0EB3
86
15
41
69
STACK2
0EB5
87
STACK3
0EB7
88
STACK4
0EB9
89
STACK5
0EBB
90
STACK6
0EBD
91
STACK7
0EBF
92
STACK8
0EC1
93
STACK9
0EC3
94
STACKA
0EC5
95
STACKB
0EC7
96
STACKC
0EC9
97
STACKD
0ECB
98
STACKE
OECD
99
STACKF
0ECF
100
END
0ED1
102
17
STMT
ADDR
CODE
STATEMENT
1
0000
8D
OE
PRGRM1
BSR
SETUP
2
0002
86
30
LOOP1
LDA
A
#’0
3
0004
8D
18
BSR
OUTPUT
4
0006
20
FA
BRA
LOOP1
5
0008
8D
06
PRGRM2
BSR
SETUP
6
OOOA
86
31
LOOP2
LDA
A
#'l
7
000C
8D
10
BSR
OUTPUT
8
000E
20
FA
BRA
LOOP2
9
0010
FE
001C
SETUP
LDX
ACIA
IQ
0013
86
13
LDA
A
#$13
11
0015
A7
00
STA
A
0,X
12
0017
86
11
LDA
A
#$11
13
0019
A7
00
STA
A
o,x
14
001B
39
RTS
15
001C
8008
ACIA
FDB
$8008
16
001 E
DE
1C
OUTPUT
LDX
ACIA
17
0020
C6
02
T1
LDA
B
#$02
18
0022
E4
00
AND
B
o,x
19
0024
27
FA
BEQ
T1
20
0026
A7
01
STA
A
1,X
21
0028
39
RTS
SYMBOL VALUE
DEFN
REFERENCES
PRGM1
0000
1
LOOP1
0002
2
4
PRGRM2 0008
5
LOOP2
OOOA
6
8
SETUP
0010
9
1 5
ACIA
001C
15
9 16
OUTPUT 001E
16
3 7
T1
0020
17
19
Program B. Test program.
107
ASTABLE MONOSTABLE
CI«.OOOOI TO lOOpE R3, C2 DETERMINE PULSE WIDTH
RI.CI DETERMINE FREQUENCY
Fig. 1. Interrupt-oscillator circuit.
Cl may be changed to vary the
interrupt rate.
The question of how often we
generate an interrupt now
arises. Suppose we were to
generate an interrupt once
every ten seconds. If each user
were printing out data, the
printing would be done in
spurts. Another problem would
be that a user might type in
data while another program
was being run, resulting in in-
put being lost. If we increased
the interrupt rate fast enough,
the output would appear
smooth and continuous. Also it
would be impossible for a per-
son to type so fast that some
data might be lost. So, it would
seem that the faster we gener-
ate interrupts, the better.
The problem, however, is that
the service routine takes a fixed
amount of time to perform its
duties. As we increase the rate
of interrupting, the percentage
of time the microprocessor is in
the service routine increases. It
is possible to generate inter-
rupts so fast that 99 percent of
the time is spent in the service
routine, meaning that only one
percent of the processing time
actually performs useful work.
Therefore, we should try to
choose an optimal interrupt
rate. I find that 100 interrupts
per second works well. You
should experiment to deter-
mine what works best for you.
You could also determine the
optimal rate mathematically;
this would require that you ex-
amine matters in more detail.
Programming Considerations
Suppose you are time-shar-
ing two or more programs at the
same time. If these programs
are in different segments of
memory, there are no problems.
Often, however, it is desirable
that programs be able to share
the same subroutines; this is
necessary for large programs.
For example, BASIC might
take up approximately 8K
bytes. If each of four users had
his own copy of BASIC, we
would need at least 32K! If all
four users could use one copy
of BASIC at the same time we
would need only 8K, resulting in
a tremendous saving in mem-
ory (of course, each user still
needs his own area to store his
program).
But wait a minute! You can-
not take any subroutine and ex-
pect it to work on a time-shared
basis. As a matter of fact, most
subroutines would not work at
all. A subroutine that is reen-
trant is needed. A reentrant
subroutine is defined as one
that may be employed by many
users at the same time (i.e., on
a time-shared BASIC). Let’s go
over some examples of reen-
trant and non-reentrant
subroutines.
Let’s say we wanted to write
a subroutine that would add the
contents of the A register to
that of the B register and store
the result in the B register. It is
also desired that the A register
not be modified when we return
from this subroutine. The sub-
routine in Program C will ac-
complish this for a single user
and will prove to be non-
reentrant.
Suppose two users call this
routine at about the same time,
and the values of the A register
for both users are $01 and $02,
respectively, upon entry into
the subroutine. User 1 enters
the subroutine and executes
the first three instructions
before an interrupt occurs.
Location TEMP will then con-
tain a value of $01.
Let us now assume that after
the interrupt, program 2 enters
the subroutine and is inter-
rupted after three instructions
have been executed. Location
TEMP now has a value of $02.
After the interrupt, user 1 will
resume execution and execute
statement 4, a load instruction.
The A register will now contain
a value of $02. We will then
return from the subroutine.
You will immediately notice
that from user 1’s point of view,
the value of the A register has
been changed from $01 to $02
upon leaving the subroutine.
This was not intended. So, we
have here an example of a
subroutine that works for one
user, but falls apart for two.
Now, let us write the same
subroutine in a different way,
as shown in Program D. This
subroutine turns out to be reen-
trant. We’ll assume the same
sequence of events as in the
previous example. User 1 will
save $01 by pushing it onto its
own stack. When user 2 enters
the subroutine, it saves $02 on
its own stack. The crucial point
here is that each program has
its own stack. Consequently,
$01 and $02 are stored in dif-
ferent locations. When each
program executes the PUL A in-
struction, it does so with
respect to its own stack. This
means that the proper values
are restored. Two or more users
can therefore use this sub-
routine at the same time!
Another example of reen-
trant programming can be
found in the Motorola M6800
Programming Manual. For ex-
ample, on pages 10-12 a reen-
trant 16-bit multiplication sub-
routine is depicted. The key
technique here is that every-
thing is first pushed onto the
stack. The TSX (T ransfer Stack
Pointer to Index) is then ex-
ecuted. All instructions that
follow are executed in the in-
dexed mode. This is equivalent
to the work area being in the
stack. Nowhere in the program
is there a label designating a
storage location.
STMT ADDR CODE
1
0000
B7
0009
2
0003
IB
3
0004
16
4
0005
B6
0009
5
0008
39
6
0009
SYMBOL VALUE
DEFN
ADD
0000
1
TEMP
0009
6
STATEMENT
ADD
STA
ABA
TAB
A
TEMP
LDA
RTS
A
TEMP
TEMP
RMB
1
REFERENCES
1 4
SAVE A-REGISTER
RESTORE A-REGISTER
Program C. A non-reentrant subroutine.
108
STMT
ADDR
CODE
STATEMENT
1
0000
36
ADD PSH
A
SAVE A-REGISTER
2
0001
IB
ABA
3
0002
16
TAB
4
0003
32
PUL
A
RESTORE A-REGISTER
5
0004
39
RTS
SYMBOL VALUE DEFN
REFERENCES
ADD
0000
1
Program D. A reentrant subroutine.
In general, writing reentrant
subroutines may be easy or dif-
ficult, depending on the type of
instruction set available. For
example, if the M6800 micro-
processor had a PSH X instruc-
tion, the task of reentrant pro-
gramming would be greatly
simplified. Other processors
have defects of their own.
Perhaps in the future someone
will design a stack-oriented
microprocessor. Reentrant pro-
gramming may then become a
trivial task. Incidentally, stack
processors have other advan-
tages than the one given.
You must be careful, though,
that the stack pointer does not
change too much from its initial
value. At the start of execution,
the stack pointers of all pro-
grams initially differ by 16. This
will change slightly throughout
the course of execution. For ex-
ample, if we were in program 1,
an interrupt might occur after
we had jumped to a subroutine.
This would cause the stack
pointer to differ by 2 from its ini-
tial value. If we nested sub-
routines too deeply, say 8 or 9,
we could change the stack
pointer so much that we’d wipe
out the stack of another pro-
gram! This problem can be
solved, however, by initially
separating the stack pointers
by more than 16.
Since the time-sharing
routine uses the stack pointer
for its own bookkeeping, you
must be careful what you do
with the stack pointer. A com-
mon technique is to use the
stack pointer to point to a list of
numbers. This will not work if
the stack pointer is pointing to,
say, the middle of a list of
numbers. It won’t work be-
cause on interrupt, the regis-
ters that are stored in the stack
will destroy some numbers in
the list. Jumping to a subrou-
tine or doing PSHES and
PULLS modify the stack pointer
but are not harmful because
the stack pointer is changed in
a way that won’t change valid
data in the stack.
Remarks
In this article, I have tried to
point out some of the essential
points that must be understood
in order to implement time-
sharing. I hope I’ve taken some
of the mystery out of it. ■
Hems how the Computer
Workshop works for you.
Our clients get the whole thing—
total systems engineering. We can
help you with evaluation and devel-
opment of requirements for data
processing and associated hard-
ware, or give you management
assistance in developing systems.
We’ll develop and design computer
systems, and well design, organize
and manage special training pro-
grams for your staff.
Here’s where to find us: offices: Baltimore 4005 seven m
Building, 4170 Wm. Penn Highway, Murraysville, Pa. 15660 (412) 327-1
Rockville, Md. 20852 (301) 468-O455/N0RTHERN VIRGINIA 5240 F
KANSAS CITY 6 East Street, Parkville, Mo. 64152 (816) 741-5055
So even if your needs are minimal
right now, we can be a big help.
Then as your needs grow, you’ll
know where to come.
The Computer Workshop
Here are some of the firms we represent:
Alpha Microsystems
Compucolor Corp.
The Digital Group
International Data Systems, Inc.
Lear Seigler
North Star Computers
Newtech Computer Systems
Sanyo Electronic, Inc.
Southwest Technical Products Corp.
Vector Graphics
Lane, Baltimore, Md. 21208 (301) 486-5350/PITTSBURGH Robar
^/WASHINGTON, D.C. AREA 1776 Plaza, 1776 E. Jefferson Street,
t Royal Road, Suite 203, Springfield, Va. 22151 (703)321-9047/
Cassette Recorder
Disaster:
Ground Loops
the problem, and a solution
Photo 1. Cable adapter made from standard parts.
Dave Waterman
834 Oak Lee Ln.
Alpine CA 92001
Dave Lien
8662 Dent Dr.
San Diego C A 92119
T he ordinary household cas-
sette recorder was not
designed with anything as ex-
otic as digital data recording in
mind. Computer experimenters
pressed the recorder into this
role. All things considered, the
device works well. However,
two problems immediately
arise— low-level ground loops,
which can badly degrade the
system’s reliability, and the
lack of a convenient means of
overriding the computer’s con-
trol of the drive motor. We’ll ad-
dress the problems separately.
Photo 2. The ungrounding adapter is inserted between the com- * he Ho Hum
puter and the recorder in the EAR or AUX lines. The standard cassette
recorder was not designed to
input audio (data or otherwise)
via its AUX or MIC jack, and an
instant later feed audio out
through the EAR jack— with all
jacks tied to a common exter-
nal ground. Many recorders do
not even have a common inter-
nal ground for these jacks and
the REM motor control jack.
Those that do usually have a
relatively high-resistance
ground. When this shaky
ground system is tied to the
computer’s common ground by
way of three separate shielded
cables (DATA-in, DATA-out and
REMOTE motor control), the
ground loops created can com-
pletely destroy the reliability of
the recording system.
A Way Around this Hummer
The standard way out of this
ground-loop problem is to
unplug either the DATA-in or
DATA-out plug from the record-
er, whichever is not in use. It
usually works but is inconve-
nient, particularly for the half-
way serious computer user who
values his time. Fortunately,
there are a couple of simple
and inexpensive solutions (un-
til more suitable recorders hit
the market at the right price).
Photo 1 shows a simple
cable adapter made from stan-
dard parts. It consists of a mini-
plug, minijack and a short (the
shorter the better) piece of un-
shielded wire. This wire is
soldered only to the “hot”
(center) connectors of both
plug and jack.
This ungrounding adapter is
inserted between the computer
and the recorder in the EAR line
or the AUX line, as shown in
Photo 2. Given the choice, it is
better to use an unbroken
shield to the AUX jack to assure
a good-quality recording. A
properly recorded tape can
always be reloaded, but a bad
tape cannot. Keep power sup-
plies and other possible
sources of interference away
from this unshielded adapter. It
works well.
The second ground-looping
solution is a variation on the
same theme, but it also solves
the annoying problem of lack of
convenient motor control. Two
jacks, one miniature (to match
110
MIC CABLE
Fig. 1.
the EAR plug) and one submini-
ature (to match the REM motor
plug), are mounted in a small
plastic case. The one shown in
Photo 3 was used to hold a
burglar-alarm panic switch. A
shielded cable is run from the
EAR jack in the box to the EAR
plug for the recorder. Note in
Fig. 1 that the shielded part of
the cable is not attached to
break the ground loop. Another
shielded cable is run from the
REM jack to the REM plug for
the recorder, but its ground in-
tegrity is maintained.
Similar switch boxes are
equipped with an SPST normal-
ly closed switch. If this is the
case with the one you select,
replace the switch with a
similar SPST switch with nor-
mally open contacts, as shown
in Fig. 1. Unshielded jumper
wires are then connected from
the switch to the subminiature
REM Jack-in-the-box (sorry
about that!). Paralleling the
REM line with the push-button
switch allows us to turn on the
motor.
We can always turn the re-
corder off with its normal STOP
button. This arrangement
allows us to turn the motor on
for purposes of rewinding tape,
advancing a cassette past the
leader or going fast forward to
find a certain spot on the tape.
Photo 4 shows this handy
auxiliary control box installed
with a Radio Shack TRS-80
Photo 3. Small case with mounted jacks.
computer system. It should
work as well with any other.
Success
Both of these solutions to
nuisance problems work well,
are inexpensive and require no
special tools or skill. Give them
a try, and see how much more
you enjoy your computer. ■
111
Glen Chamock
864 Palomar Way
Oxnard CA 93030
A Different Search Technique
don’t just try it— benchmark it
G ood things can come in
small packages. This pro-
gramming trick is so simple it
can easily retrofit to existing
programs; yet, it can substan-
tially reduce the time needed to
search a table.
The traditional method of
searching a table is shown in
Fig. 1. First, a loop index is ini-
tialized. Then a loop is execut-
ed, comparing the table ele-
ment with the search argument
and incrementing the loop in-
dex until either a match is
found or the table is exhausted.
When the loop is exited, the
loop index points either to the
location of the matching table
element or, if no match was
found, to the last table element
plus one.
The new method dimensions
one extra place at the end of
the table for a “dummy” value.
To search the table, first move
the search value into this dum-
my location at the end of the
table; then initialize the loop in-
dex and begin looping through
i
i
♦
Fig. 1. Traditional table-search-
ing method.
the table. This time, however,
only search for a match and in-
crement the loop index within
the loop. You don’t need to test
for the end of the table ... if
you haven’t found a match by
then, you will on the last table
entry because you’ve already
moved the search argument in-
to this last entry. Thus, you
save one comparison for each
table entry searched (see Fig. 2).
Depending on the language
and the way the computer im-
plements subscripts, this trick
can save as much as half the
the time needed for the search.
That’s pretty good for such a
small change!
I learned this programming
trick from the advertising
brochure of Software Consult-
ing Services of Allentown PA.
Further details may be found in
The Art of Computer Program-
ming , Vol. 3, “Sorting and
Searching,” by Donald E.
Knuth.B
i
t
Fig. 2. A different search tech-
nique.
IDS internat,onal data systems ’ ,nc -
400 North Washington Street, Suite 200, Fails Church, Virginia 22046 U.S.A.
Telephone (703) 536-7373
Terms: Payment with order shipped
prepaid. Shipping added to COD.
Minimum order $10.00. Master
Charge accepted.
BIG SAVINGS ON PRIME COMPONENTS DUE TO MANUFACTURERS LARGE QUANTITY DISCOUNTS.
LOW POWER SHOTTKEY
74LS174
1.25
BAUD RATE GENERATOR
74LS175
1.25
74LSOO
22
74LS192
2.45
MCI 4411
9.90
74LS02
24
74LS193
2.45
74LS04
24
74LS08
28
TTL
600MHz PRESCALER
74LS1 1
.35
74LS13
.69
7400
.15
11C90
19.00
74LS30
.26
7406
.29
74LS32
.34
74121
.35
74LS74
.49
74150
.85
PRIME INTEL
74LS85
1.99
74LS90
.89
SHOTTKEY
8085 CPU
29.00
74LS92
.89
8251 USART
10.50
74LS125
.63
74S00
.40
8212 PIO
4.00
74 LSI 26
.63
74S30
.40
8155
29.00
74LS151
.95
74S132
1.10
8253 PIT
24.00
74LS155
125
74S196
3.80
8255 PPI
22.00
112
A22
AP DIP Jumpers are the low cost high
quality solution to bussing between PC
boards, mother boards, backplanes and
more. Available in 14, 16, 24 and 40 pin
single-ended or double-ended assemblies,
in standard lengths of 6, 12, 18, 24 and 36
inches. Each assembly has molded-on strain relief
and line-by-line probeability. Contact material is non-
corrosive alloy 770. Dielectric is 94 V-0 rated, and
cable is stranded rainbow.
Order from your A P distributor today.
Our distributor list is growing daily. For
the name of the distributor nearest you
call Toll-Free 800-321-9668.
Send for our complete AP catalog The
Faster & Easier Book.
AP PRODUCTS INCORPORATED
Box 1 10 • 72 Corwin Drive, Painesville OH 44077
(216)354-2101 TWX: 810-425-2250
A P DIP Jumpers for Faster & Easier
connections.
The NewMicro Executive II.
It handles big jobs at a price
that makes sense to small
businesses.
Finally there's a computer
system that can run your of-
fice without eating up your
profits. For less than $15,000
(or a just-as-
reasonable leas-
ing arrange-
ment), you
can put Micro Executive II to
work handling Accounts
Payable and Receivable,
Payroll, Inventory and
Financial Reporting. (And it
can be operated by people
without extensive accounting
backgrounds.)
What's more, Micro
Executive II includes the
best programs system to be
found on any micro computer.
Besides a complete
business
system— v
programs
for special use, such as in
medical, dental and insur-
ance offices.
It also has a word process-
ing system that provides you
with a full range of editing and
editorial services, including
legal documents and
contracts, catalogues, letters
to your customer.
Doesn't it make sense?
Gorrqxiter
Workshop
Offering Microcomputer BusinessSystems
Offices: BALTIMORE 4005 Seven Mile Lone, Baltimore, Md. 21208 (301 ) 486-5350/KANSAS CITY 6 East Street, Parkville, Mo. 64152(816) 741-5055/
PITTSBURGH Robor Building. 4170 Wm. Penn Highway, Murraysville, Pa. 15660 (412) 327 045 5 /WASHINGTON, D C. AREA 1776 Plaza 1776 E
Jefferson Street, Rockville, Md. 20852 (301) 468-0455/NORTHERN VIRGINIA 5240 Port Royal Road, Suite 203, Springfield Va 2215 1 (703) 321-9047
#• / iff/ d att&ifi&d
Kilobaud classified advertisements are intended for use by those
individuals desiring to buy, sell, or trade used computer equipment
or software. No commercial ads are accepted.
Two sizes of ads are available. The $5 box allows five lines of about
22 characters each, including spaces and punctuation. The $10 box
provides ten lines of type — again, each line is about 22 characters.
Minimize capital letters, as they use twice the space of small
characters. Payment is required in advance with ad copy. We cannot
bill, or accept credit. Oversize ads are not accepted. Each subscriber
is limited to two (2) identical ads in any given issue.
Advertising text and payment must reach us 60 days in advance of
publication. For example, advertising copy for the March issue
(mailed in February) must be in our hands on January 1. The
publisher reserves the right to refuse a questionable or not
applicable advertisement. Mail advertisements to: KILOBAUD
CLASSIFIED, Kilobaud, Peterborough, NH 03458. Do not include
any other material with your ad, as it may be delayed.
Stock Market, Commodity Time Series
analysis programs. Exchange with others
who have worked in this area on computers.
Bob Lemeer, PO Box 29, Brooklyn NY
11230.
PET Owners: Perfect your system before
trying Las Vegas. Blackjack, baccarat, rou-
lette, craps. Full LV casino rules & options
give maximum reality. Full screen graphics.
All on 1 cassette. $25. CMS, 431 Monte
Vista, Dallas TX 75223.
Augat w/w boards, 24-16 pin cap. or any
comb, of other sizes. 36 dual w/w edge conn.
$8, no ICs, $9 w/ICs $10 w/ICs & edge conn,
ppd. in U.S. R. Savilonis, 334 Columbia St.
Cambridge M A 02141, (617) 492-2095 5-7
PM EST.
HP-97 prog/printing calc. Top-of-line.
Orig. $750, asking $475. P. Davis (201)
885-1220 (days).
Sophisticated 2-player football simulation
on cassette for Radio Shack TRS-80 Level I
BASIC. Uses graphics to draw field, move
ball, etc. $12. Dr. S. Harter, Box 17222,
Tampa FL 33682.
TRS-80 Sort: in BASIC 1, 4K min. Sorts recs
in memory on 1 or 2 numeric fields in asc. or
desc. seq. Input from tape, k.b. or both.
Output to tape or video. Can be subroutine.
$10 on tape w/doc. To: Software, Box 6153,
Syracuse NY 13217.
PET Software: Game and demo programs
written in BASIC for 4K PET. Includes
blackjack, children’s math, biorhythm and
15 others. On cassette, ready to load and
run, plus written descrip. $15.95. R.
Schmidt, 14 Tinker Rd„ Nashua NH 03060.
PET users need info? Getting started with
your PET workbook, $4. Workbook has
helpful info, sample programs, exercises and
fundamental features of PET BASIC. Send
SASE for details. D. Smith, Box 921, Los
Alamos NM 87544.
Heathkit H9 video terminal assembled,
$400. Kilobaud issues 1 to 15, $25. Byte
issues 3 (Nov 75) to 16 (Dec 76) in binder,
$35. Byte Jan 77 to date, $15. First money
order. T. Halldorson, 44 Ctr. Grove Rd.,
Apt D-13, Randolph NJ 07801.
Use Kilobaud Classified
„ NEW
Products
Contest !
Another month has gone
by, and the votes have been
counted. The article winner
for the wintry month of Febru-
ary is Dr. Mark Boyd, author of
“Interfacing Tips” on page 72.
Choice-of-a-book-from-the-
Book Nook winner is Larry
Nelson of Marion IN.
To both Mark and Larry, we
offer congratulations and
best wishes.
And to all of our readers
who are responding enthusi-
astically with their votes, we
also offer congratulations,
best wishes and good
reading.
Keep voting!
(from page 15)
passage. Kinged pieces are iden-
tified on the display and messages
appear at the right of the board
relating to each move.
Included with Video Checkers
is a 9-page instruction book. The
game’s author has invited pur-
chasers of the cassette to add en-
hancements to the program.
Therefore, an entire program list-
ing is included.
RS-232 ADApter for KIM
Connecting an RS-232 ter-
minal to KIM is easy with ADA.
This ADApter converts KIM’s
20-mA current-loop port to an
RS-232 port. ADA does not af-
fect the baud rate and uses stan-
dard power supplies. The unit
CmC’s ADA.
comes complete with instruc-
tions, is assembled and tested and
measures 3 x 3Vi x 1 inches. KIM
is isolated from the RS-232 device
by optoisolators.
ADA sells for $24.50 with
drilled, plated-through solder
pads for all connections, or for
$29.50 with barrier strips and
screw terminals.
CONNECTICUT microCOM-
PUTER, 150 Pocono Road,
Brookfield, CT 06804.
This Is the Electric Pencil
The Electric Pencil is a charac-
ter-oriented word-processing sys-
tem. This means that text is en-
tered as a continuous string of
characters and manipulated as
such, allowing the user freedom
and ease in the movement and
handling of text. Since lines are
not delineated, any number of
characters, words, lines or para-
graphs may be inserted or deleted
anywhere in the text. The typing
of carriage returns as well as word
hyphenation is not required since
each line of text is formatted
automatically. Whenever text is
inserted or deleted, existing text is
pushed down or pulled up in a
wraparound fashion. Everything
appears on the video display
screen as it occurs, which
eliminates any guesswork.
When text is printed. The Elec-
tric Pencil automatically inserts
carriage returns where they are
needed. Numerous combinations
of line length, line spacing and
page spacing allow for any form
to be handled. Pages may be
numbered as well as titled.
Minimum system hardware
requirements: (Stock system
— other versions available by
special order) SOL-20/SOLOS
(or equivalent: system w/ VDM-1
and CUTS cassette), 8K of mem-
ory starting at location 0000,
Printer (TTY, Selectric, etc.),
Video Display Monitor, Cassette
recorder.
The Electric Pencil is available
from stock at: Interactive Com-
puters, 7646 1/2 Dashwood,
Houston TX 77036.
Microprocessor Protection
Lightning and heavy-duty elec-
trical equipment often create
power-line surges and transients.
These can cause extensive dam-
age to valuable microprocessors
and peripherals.
Electronic Specialists announc-
es a line-cord transient sup-
pressor that will absorb repeated
power surges to protect delicate
equipment.
Available in 2-prong plug/
socket ($11.50) or 3-prong plug/
socket ($14.50), these units are
also available with integral
power-line hash filtering.
Electronic Specialists, Box
122, Natick MA 01760.
Electronic Specialists’ transient suppressor.
114
WHY YOU
SHOULD
MAKE A
CORPORATE
CONTRIBU-
TION TO
THE AD
COUNCIL
The Advertising Council is the biggest
advertiser in the world. Last year, with
the cooperation of all media, the Coun-
cil placed almost six hundred million
dollars of public service advertising.
Yet its total operating expense budget
was only $914,683, which makes its
advertising programs one of America’s
greatest bargains ... for every $1 cash
outlay the Council is generating over
$600 of advertising.
U.S. business and associated groups
contributed the dollars the Ad Council
needs to create and manage this
remarkable program. Advertisers, ad-
vertising agencies, and the media
contributed the space and time.
Your company can play a role. If you
believe in supporting public service
efforts to help meet the challenges
which face our nation today, then your
company can do as many hundreds of
others— large and small— have done.
You can make a tax-deductible con-
tribution to the Advertising Council.
At the very least you can, quite easily,
find out more about how the Council
works and what it does. Simply write to:
Robert P. Keim, President, The Adver-
tising Council, Inc., 825 Third Avenue,
New York, New York 10022.
Vj^yi I A Public Service of This Magazine
(Sukll & The Advertising Council.
,\^v°
<o v*
VIDEO
TERMINAL
Now, a completely self-contained video terminal card for less than SI 50.00.
Requires only an ASCII Keyboard and TV set to become a complete interactive
terminal for connection to your computers serial 10 port. Two units available,
common features are: single 5V supply, crystal controlled sync and baud rates
(to 9600 baud), computer and keyboard operated cursor control, parity error
and control, power on initialization, forward spaces, line feed, rev. line feeds,
home, return cursor, and clear to end of line. Power requirements are 5V at
900ma. output std. IV p-p video and serial TTL level data.
Features:
TH3216
TH6416
Display
32 characters
by 1 6 lines
2 pages
64 characters
by 16 lines
scrolling
Characters
Upper case ASCII
Upper/lower <
optional
Baud Rates
300-9600
110-9600
Controls
Read to/from
memory
Scroll up or
down
Price (kit)
$149.95
$189.95
Above prices include all 1C sockets
OPTIONS:
Power supply (mounts on board) $14.95
Video/RF Modulator, VD-1 6.95
Lower case option (TH6416 only) 10.00
Assembled, tested units, add 60.00
SliEEKflEU
"TH 6416 shown above"
Frequency Counter
$ 79 . 95 KIT
You ve requested it. and now it s Here' The CT-50 Fre-
quency Counter Kit has more features than countars selling
for twice the price Measuring frequency is now as easy as
pushing a button, the CT-50 will automatically place ' h e
decimal point in all modes, giving you quick, reliable read-
ings Want to use the CT-50 motxle - ’ No problem it runs
equally as well on 12 VDC as it does on HO VAC Want
super accuracy'’ The CT-50 uses the popular TV color burst
freq ol 3 579545 MHz tor time base Tap oft a color TV with
Our adapter and get ultra accuracy — 001 ppm 1 The CT-50
offers professional quality at the unheard ol price of S79 95
Order yours todayl
SPECIFICATIONS
Sensitivity less man 25MV
Frequency range 5Hz to 60MHz typically 65MHz
Gate lime 1 second. 1 10 second with automatic decimal
point positioning on both direct and prescale
O splay 8 digit red LED 4 height
Accurac, 2 ppm. 00f ppm with TV time base 1
Input BNC I meg ohm direct 50 ohm with prescale option
Power 110 VAC 5 wans oi 12 VDC a.^Amp
Size Approx 6x4 x 2 . high quality aluminum case
PRICES
CT-50, 60MHz Counter Kit $79.95
CT-50WT, 60 MHz counter, wired and tested $159.95
CT-600, 600 MHz prescaler option for CT-50, add . . $29.95
VIDEO TO RF
MODULATOR
Convert any TV set to a
video monitor. Super stable
circuit is glitch-free, tunable
over channels 4-6. Runs on
5-1 5V. Recommended by
many computer manufactur-
ers. Std. video input. Com-
plete kit, VO-1 S6.95
741 OP-AMP
MINI DIP 10/S2 00
LINEAR
REGS
TRANSISTORS
555
.50
309K
.99
NPN 2N3904 type
10/S1.00
556
.75
340K-12
.99
PNP 2N3906 type
10/SI .00
566
1.49
7805
.99
NPN Power Tab 40W
3/SI .00
567
1.49
7812
99
PNP Power Tab 40W
3/SI .00
324
1.49
7815
.99
FET MPF-102 type
3/S2.00
1458
.49
78MG
1.50
UJT 2N2646 type
3/S2.00
380
1.49
723
.49
2N3055 NPN Power
75
RS232/TTL
TTL/RS232
Converter kit
Complete kit S7.95
1C SOCKETS
8 pm
low profile
5/S 1.00
14 pin
low profile
5/S1.00
16 pm
low profile
5/SI. 00
40 pin
low profile
2/SI. 00
14 pin
wire wrap
3/Si 00
lYiivtsij sleiafrniiBS
Box 4072 K ROCHESTER NY 14610 (716) 271^487
Satisfaction guatan
t**a or monay refund
® ad. Order, under S10
I 76c COO- edd
00 NY eed 7%
order M OO »■<•* »**
MINI-KITS
FM WIRELESS MIKE KIT
Transmit up to 300’ to any FM
radio. Sensitive mike input re-
quires dynamic, crystal or ceram-
ic mike. Runs on 3 to 9 volts.
FM 1 S2.95
TONE DECODER KIT
A complete tone decoder on a Single PC
Board Features 400 lo 5000 Hz adjustable
frequency range, voliage regulation 567 1C
Ueeful Or touch -'one decoding tone b'-rsi
detection. FSK demod signaling and many
other uses Use 7 lor 12 Dutton louchione de
coding Runs on 5 to 12 volts
Complete Kit. TD 1 >4 96
LED
BLINKY KIT
A great attention gel
ter which alter naiely
hashes 2 iumoo LEDs
Use to' name badges,
buttons or warning
lype panel lights
Complete Kit. BL 1 .
ft
SUPER-SNOOP AMPLIFIER
A Super -sen si live amplihei which will pick
up a pin drop at 15 leet 1 Great tor monitoring
baby's room or ss s genera! purpose lesi
amplifier Full 2 watts ol output, runs on 6 lo
1 2 vo Its. uses any type ot mike Requires 8-45
Complete Kit. BN-8
>4 96
MUSIC LIGHTS KIT
See music come alive 1 3 different lights
dicker with music or voice One light tor
lows, one for the mid-range and one tor the
highs Each channel individually adjustable,
and drives up to 300 watts Great tor parties
band music, nite clubs and more
Complete Kit. Ml-1 $7 96
SIREN KIT
Produces upward and downward wail char-
acteristic ot police siren 20Onw audio out-
put. runs on 3-9 volts, uses 8-45 ohm
Complete Kit SM-3 $2.96
POWER SUPPLY KIT
Complete triple regulated power supply pro-
vides variable ±15 volts at ?00ma and +5
volts at i amp SOmv load regulation, good
filtering and small size Kit less translormeis
Requires 6-BV at t amp and 18 to 30VCT
Complete Kit. PS-3LT $6.96
115
NOW HEAR THIS ! ! !
HEX DISPLAYS ARE HERE!
• 4 bits in dispays 0 through F
• Built in latch and leading zero blanking
• Use 6 modules to display hex address and data
• Connect 2 modules to an output port to display
registers, data, etc.
• End confusing strings of discrete LEDs
7 seg. 1/3 in. display - • • $6.80
7 seg. A in. display - - - $7.50
NORTH STAR SOFTWARE
• Variety package #1 - - inventory, hex and dec.
translators, memory fill, electronic Design, memory
test, games, and more complete package on one
double-sided diskette - - $20
• Additional double-sided diskettes - - $5
Add SI shipping and handling for orders under $20
MC or VISA welcome 24 hours
MIDWEST DIGITAL
863 Wood Ave. Wichita, KS 67212
316-722-1711 M42
The Southern California Swapfest is being held in Santa Bar-
bara on July 1. John Craig says, “We’re going to have an old-fash-
ioned, down-to-earth, good-time swap meet for computer hobby-
ists, electronics experimenters and hams from all over California.”
Atlanta GA
The 16th Annual Convention of the Association for Educational
Data Systems will be held in Atlanta GA, May 15-19, 1978. For fur-
ther information, contact: Dr. James E. Eisele, Office of Com-
puting Activities, University of Georgia, Athens GA 30602.
Washington DC
Amateur Computing 78 microcomputer festival will be held July
22-23 at the Sheraton National Motor Hotel, Arlington VA.
INTRODUCING
THE OE 1000
VIDEO TERMINAL
The OE 1000 is a low cost terminal for use with
any computer having serial ASCII capability.
Just add a modified TV or video monitor. The
following features can be found on the OE 1000:
• 16 lines x 64 characters
• 128 characters, including upper and lower
case
• Has full x-y cursor control
• Keyboard operates in either upper/lower case
or TTY mode
• Packaged in high impact plastic case
• RS 232 or 20 ma current loop
Price for the OE 1000 is
$275.00 kit form • $350.00 assembled
Call or write today. MC or BAC accepted
OTTO ELECTRONICS
P.O. Box 3066 Princeton. N.J. 08540
(609)448-9165
116
"PIXE-PLEXER "
Ends those SPECIAL Video Monitor PROBLEMS!
MODEL
PXP-a500
$24.50
kit
Pixe-Plexer is an 1C type modulator-RF oscillator for
interfacing audio, color and regular monochrome video
signals from computers, TV cameras, VTR's, games, etc.,
for display on any regular TV set via the antenna
terminals. Power requirements: 15V @50 ma max.
Operates on ch. 2-6 tunable. Includes 3.58 Mhz color
subcarrier and 4.5 Mhz audio subcarrier with varactor
modulator. R-Y and B-Y inputs. Analog or digital inputs.
Complete with 1C data sheet and instructions. Numerous
circuit variations from deluxe to simple.
For all other video display needs, don't forget our
popular "Pixe-Verter" Mod. RF kit Model PXV-2A.
$8.50.
Available from your local dealer or factory-direct. Phone
or write for additional assistance. Dial 402-987-3771.
-W-
-t/H
13-K Broadway ATV Research Dakota City, NE.
A43 68731
Datapoint CRT Terminals
Fully-Assembled — Guaranteed
#3360 $ 649.50
• Add $15 packing. refurbished
• Shipped FOB Washington. D C. Terms check, M.O.
or charge.
• 90-day guarantee • Scrolling version $695.00
Model 3360 speeds from 300-4800 Baud, numeric
keypad, cursor controls. Edit. Block-T ransmit, search
modes. ASCII Keyboard with codeable options.
Green phosphor. 24 80 ch lines, addressable cursor;
RS-232C serial interface; other speeds available.
Manual $10; cable kit $9.95. Datashare/IBM-2260
compatible version $1 .1 00.00 • Model 3000 $825.00.
M-33 ASR Teletype $895. KSR $725; All M-28. 35
components available, also Modems, readers.
QUANTITY DISCOUNTS AVAILABLE. Leasing, ser-
vice at low prices.
TELECOMMUNICATIONS SERVICES CO.
Box 4117, Alexandria, Va. 22303 T26
703-683-4019 / TLX 89-623
Complete retail & service computer
outlet in
LONG ISLAND
FREE CONSUL TA TION ON A LL
YOUR NEEDS
In stock — IMSAI, Polymorphic, TDL,
Digital Systems, Tarbell, Soroc,
Large variety of memories,
interfaces, software, books & magazines
PRINTERS • FLOPPIES • CRTs
Custom designed systems for
personal, educational & business use
Free Newsletter — Systems
Demonstration — Best Service
Computer ITlicrosy stems
1 31 1 Northern Blvd.
Manhasset NY 11030 Q 44
(516) 627-3640
Open Tues-Sat: Wed-Sat 10-6;Tues 10-9
“REMark" is published quarterly
Copyright © 1978, Heath User's Group
Two Heathkit User Organizations??
BULK RATE
U.S. Postage
PAID
Heath Company
You may have noticed that there are two entirely
separate organizations for Heathkit computer users.
HUG is the official Heath sponsored organization.
The second organization publishes a periodic news
letter called BUSS. This second organization is not
affiliated with Heath Company in any way. Neither is
it approved, sanctioned, or recommended. Heath
Company bears no responsibility for the material it
publishes or the advice it gives. The official, inside
word will always come through HUG. If you want the
latest word on new products, software updates, and
other juicy news, keep reading REMark.
Tor
1 2*
first
CLASS
BUSSES*
325 Penney (vatu' a AvSE'.
Was P. C. 2 0003^"^
The Seven M /ssve oi BUSS
CdtntOui A hot The- Sa/tJ£~[f/n£
AsTAe F/rst Issue erf RWA
Reader Service # F4
PRINTED CIRCUIT
BOARDS
(BARE BOARDS ONLY)
29.95 EA.
YOUR CHOICE OF
S-100 BUS
8K WW EXT 4K ROM
OR
6800 BUS
8K WW EXT
We also stock components
for all above boards.
Guaranteed, if not completely
satisfied return for refund.
We back everything we sell.
SEND SASE FOR OUR CATALOG TO:
BARNES ELECTRONICS
P.O. BOX 673
OAK RIDGE, TN. 37830
PAYMENT TERMS CASH WITH ORDERS.
ALL ORDERS F.O.B. OAK RIDGE, TN.
TN. RES. ADD 4Vi%. ADD 52.00 PER BOARD
FOR POSTAGE AND HANDLING.
(WE ACCEPT MASTER CHARGE AND B.A.C.)
C^e.e.ioC
MINI DIP SLIDE SWITCH
9 Switches — $1.75
8 Switches— 1.65
7 Switches— 1.55
Write for our FREE flyer. Contains
hundreds of different items:
• 7400 TTL Series I.C/s, Linears,
Voltage Regulators
• Used Test Equipment
• Continental Spec. Sockets &
BusStrips
• Datak Corp's Complete Line of
Photo Etch Supplies
• I.C. Sockets, Ribbon Cable
• Capacitors, LED's, etc., etc. R18
R.W. ELECTRONICS, INC.
3203 North Western Avenue
Chicago, Illinois 60618
312-248-2480
VIDEO TERMINAL
M0DEL795 VIDEO DISPLAY TERMINAL
With 7x10" magnetically-
deflected CRT. With inter-
nal DC-to-DC converter,
10 KV supply, horiz. and
vert, deflection amps;
internal muffin fan. Con-
trols for intensity, vert,
and horiz. gain/position.
Requires 115 VAC 60 Hz.
Less power cord. 15 Vix
17V2x21";Sh.Wt: 100 lbs.
Used. Reparable $59.50
KRS-33 TELETYPEWRITER and
KEYBOARD - 100 WPM, 600 operations per
sec. Utilizes ASCII code with parity bit; 8 ’/ 2 " sprocket-
ted paper. 115 VAC 60 Hz. QVixWAxW/i" : Sh. Wt:
50 lbs. Used, operational, but may require some
adjustment. Less cover, $250. With cover, $300.
Write for New 78 Catalog of Government and
Commercial Electronic Surplus
All prices F.O.B. Lima, Ohio
Items over 50 lbs. shipped motor freight collect.
Address: Dept. K • Phone: 419/227-6573 pi
FAIR RADIO SALES
1016 E. EUREKA - Box 1105 • LIMA, OHIO *45802
I
Its
ytes
ooks.
argams
M17
\M
SAI 8080
«L
MHK
MARKETLINE SYSTEMS, Inc.
2337 Philmont Ave.
Huntingdon Valley, Pa. 19006
215/947-6670 800/523-5355
2708/16
EPROM PROGRAMMER
for the F-8, 6800, 8080, 1802,
or KIM- 1 microcomputers.
Features:
• RAM starting address,
PROM starting address,
number of bytes to be
programmed can be easi-
ly specified.
• Includes software for
verifying programming.
• Hardware requires only
1 Vi I/O ports to interface
to any microcomputer.
• Low insertion force
programming socket.
Assembled and tested w/software $59.95, Kit $49.95.
Kit w/out software but w/software instructions
$33.00. We ship 3 days ARO for M.O. or C.O.D. After
1 PM 8 04-973- 548Z
Optimal Technology, Inc.
Blue Wood 127
Earlysville, VA 22936 010
117
mm
MICRODESK
Overseas agents and distributors
are waiting to sell your products
in foreign markets.
The perfect home
for a personal or business computer
48” x 24” x 26”, Shelves 23” x 23”, One Adjustable
MICRODESK $96.50 113 lbs. F.O.B. Wichita
COMPUTER SYSTEMS DESIGN
1611 E. Central Wichita, Kansas 67214
DEALER INQUIRIES INVITED
Find out how to find them. Fast.
You’ve got a great prod-
uct. With great domestic
sales. And now you want
your share of the annual
$107 billion U.S. export
market. But who’s going
to handle that product
thousands of miles from
home? The U.S. Com-
merce Department can tell
you. Because through our
Agent/Distributor Ser-
vice, we’ll match your
product with up to six
firms that are ready, will-
ing and able to represent
you in the countries you
choose. To learn more,
write Secretary of Com-
merce, BIC 11C, Wash-
ington, D.C. 20230.
A Public Service of This Magazine & The Advertising Council
The 1978 Atlanta HamFestival
and
Georgia State ARRL Convention
June 3*4) 1978
Downtown Atlanta Marriott Hotel
• GIANT covered Fleamarket/Swapshop! • 140 Major Exhibits!
• More than 50 Forums/Meetings! • Special MICROPROCESSOR Section!
• FCC Exams! • Programs for Ladies & Children!
• Parking for thousands of cars! • Activities Galore!
Registration: $3 per person IN ADVANCE, $4 at the door
Ladies & Children FREE!
If you do not receive a Preregistration Packet by May 1st, write:
Atlanta HamFestival 1978
100 Woodlawn Drive
Marietta, Georgia 30067
or call Area 404/971-HAMS anytime day or night (PLEASE DO NOT CALL BEFORE MAY 1st)
Hotel Rate: $26 per day single OR double!
Write for Hotel Reservations to:
Marriott Hotel • Courtland at International Blvd. • Atlanta, GA 30303
or phone: Area 404/659-6500 and hurry, hurry, hurry!
THE BEST HAMFEST IN THE WORLD!
118
VERY AFFORDABLE
YET VERSATILE
INTEL 2716
TMS 2716
2708, 2704
$94.50
EPROM PROGRAMMER
• Our new DS 2716 EPROM Programmer connects to V/z parallel I/O
port [PIA] and 5V power supply or plugs directly Into 2716 [2708,
2704] EPROM socket on any computer via 48” flat cable
• No other wires to hook to interrupt request, wait line or external power
supply
• All control sequences, voltage and timing generated on programmer
board
• Unique handshake control assure error free programming - latched
data verified before programming pulse is applied
• Your computer does data transfer and sequence initialization over the
address lines [read only socket] or data bus [I/O port]
• On board high efficiency noiseless DC to DC converter, or 30 V minia-
ture battery for marginal power supply [good for 100 programming!]
• Programs Intel 2716, TMS 2716, 2708, 2704 [switch selectable]
• Program and Read mode [switch selectable], Verify [software select-
able]
9 Simple software instructions included for your convenience
DS 2716-K KIT [no enclosure and ZIFS] $ 94.50
DSENCL Enclosure [7y2”x4V2 M x 2”] $ 8.75
DS ZIFS Zero Insertion Force Socket $ 5.25
DS 2716-A Assembled [ENCL & ZIFS Included] $147.00
ELEN PRODUCTS, Inc
Digital Service Delivery:
P.0. BOX 1233 Kit - 7 days ARO
Akron, Oh 44309 Assembled - 15 days
1
sgSgSr^i
*»os' T,v a
R22
(209) 733-9288
8474 Ave. 296 • Visalia, CA 93277
We accept BankAmericard/Visa
and Master Charge
► Power supply included
► Rack mounted
versions from $210
Fan, line cord, fuse,
EMI filter, power switch
Desk top versions from $240
Cutouts available for most
drives: standard and mini
Desk top
model
$240
(Drive not
included)
Write or call for a copy of our
detailed brochure which includes
our application note BUILDING
CHEAP COMPUTERS.
INrEGWND
TELETYPE' MODEL 33 ASR
COMPUTER I/O COMPLETE WITH
1 Tape Punch • Ready to Go
> Tape Reader • Line/Local Wired
• Guaranteed 30 Days
$840
oo
WV fcIA
\ ACO
INCLUDING PACKING
F.O.B. N.J. FACTORY
COMPLETE COVER READY TO INSTALL .... $ 35.00
COPY HOLDER (182036) $14.00
TAPE UNWINDER (12” NEW) $ 55.00
WIND UP PAPER TAPE WINDER (1") $ 52.00
ELECTRIC PAPER WINDER (LPW 300 NEW) . $ 50.00
DEC TYPE READER RUNCARO $ 45.00
ACOUSTIC COUPLER
OMNITEK 501 A (NEW) $160.00
READER POWER PACK (182134) $ 35.00
EIA RS-232 INTERFACE $ 55.00
ACOUSTIC COVERS with FAN (NEW) $258.00
If its for TELETYPE 0 We Have It.
If you don't see what you need, CALL or WRITE!
TELETYPEWRITER COMMUNICATIONS SPECIALISTS
550 Springfield Avenue • Berkeley Heights, N. J. 07922 T13
(201) 464-5310 • TWX: 710 98&3016 • TELEX: 13 6479
SUBSIDIAR Y OF VAN'S W2DL T ELECTRONICS.
119
Same day shipment. First line parts only. Factory
tested. Guaranteed money back. Quality IC’s and
other components at factory prices.
INTEGRATED CIRCUITS
7400TTL
7400N
7402N
7409N 23
7410N .17
7414M .63
7420N 17
7422M 1.39
7430M .20
7442N 50
7445N .69
7447N .60
7448N 69
7450N .17
7474N .29
7475N 49
7485N . .88
7489M 2 00
7490N .43
7492N .43
7493N 43
7495N 69
74100N 90
74107N .29
74121 N 34
74123N 59
74125N 39
7414SN 69
74150*4 9$
74151 N 69
74154N 1.00
741S7N 69
74 1 61 N .87
74162N .87
74163N 87
74174N .96
74175*4 90
74190N 1.15
74192N .87
74193N 85
74221N 155
74298N 165
74365N 66
74386N 66
74367N .66
741S00 TTL
74LSQ0N 25
74LS02N .25
74LS04M 25
74LS05H .25
74LS06N 25
74LS10N 25
74LS13N 40
74LS14N .90
74USZ0N 25
74LS22N 2S
74LS28N 41
74LS30N 25
74LS33N .39
74LS38N .30
74LS74N 35
74LS/SN .47
74LS90N 51
74LS93N .51
74LS95N 1 .89
74IS107N 35
74LS112N 35
74LS113N .35
74LS132N .72
74LS136N .35
74LS151N .67
74LS155N .67
74LS157N .67
74LS162N .91
74LS163M .91
74LS174N .95
74LS190N 1.06
74LS221N 195
74LS256N .67
74LS367N 89
LINEAR
CA3045 .90
CA3046 .67
CA3081 1.80
CA3062 1.90
CA3089 2.95
LM301AN7AH 35
LM305H .67
LM307N .35
LM308N .89
LM309H 115
LU399K 95
LM311H/N .90
LM317T/K 2 92
LM318 1.35
IM32W-5 1.20
LM323K-5 6.95
LM320K-12 1.35
LM320K 15 135
IM32CT-5 1.60
LM329T-8 160
LM320T-12 1.50
LM320T-15 1.60
LM324N 1.15
LM339N 155
LM340K-5 1.10
LM340K-8 1.10
LM340K-12 1.10
U4340K-15 1.10
LM340K-24 1.10
IM340T-5 1.10
LM340T-8 1 10
LM340T-12 1.10
LM340T-15 1.10
LM340T-1B 1.10
LM340T-24 1.10
LM343H 4 50
LM370 1.15
LM377 4.50
LM379 5.00
LM380N 100
LM381 ISO
LM3S2 ISO
LM703H .40
LM709H .28
LM723H/N 50
LM733N .67
LM741CH .35
LM741N 25
LM747H/N 82
LM746N .35
LM1303N .82
LM13G4 110
LM1305 1 27
LM1307 2.00
LM1310 2.75
LM1458 ? 47
LM1800 1.75
LM1812 7.50
LM1889 3.00
IM2111 1.75
LM2902 1.50
LM3900N .60
LM3905 1.75
LM3909N -61
MC1458V 50
NE540L 2 89
NESSON 65
NE555V .43
NE556A .79
NE565A 100
NE566V 115
NE567V 1.20
78L05 SO
78 LOB SO
79105 70
78M05 .85
75108 1.75
75491CN .50
75492CN 55
75494CN .89
A to 0 CONVERTER
8038B 4.50
8700CJ 1395
8701CN 22 00
8750CJ 13.95
10130 9 95
940DCJV/F 7.40
ICL7103 9.50
CMOS
C034001 .50
C04000 16
C04001 .21
C04002 .21
004006 1.10
C04007 .21
C04008 21
C04009 .39
C04010 39
CD401 1 21
C04012 .21
CO4013 .36
CO4014 65
C04015 66
C04016 35
CD4017 .94
CD401B .94
CD4019 21
C04020 102
CD402I 1.02
CD4022 Be
CD4023 .21
CD4024 75
.21
CD4026 1 51
CD4027 36
CD4028 .79
CD4029 1 02
CD4M0 21
C 04 035 1 02
C 04040 1.02
C 04042 .71
C 04043 .63
C 04044 .63
CD4046 1.67 ;
CD4049 .36
C 04050 .36
CD4051 1.13
CD4060 1.42
C04066 .71
C 04068 .40
CD4069 .40
C 04070 40
CD4071 .21
C04072 .21
CD4073 .21
C 04075 .21
C04076 1.75
C 040 78 .40 1
C 04 061 .21
CD4082 .21
CD4116 .47
CO 4490 5 50
C04507 1.00
CO4508 4 25
C04510 1.02
C04511 .94
C04515 2 52
C04515 1.10
C0451B 1.02
C04520 1.02
C04527 1.51
CD4528 .79
C04553 5 75
CD4566 2 25
CD4583 4.50
C04585 1 10
CD4C192 3 30
74COO .28
74C04 .33
74C10 .28
74C14 2.10
74C20 28
74C30 .28
74C48 2.95
74C74 75
74C76 1.40
74C90 1 15
74CS3 140
74C154 3 00
74C160 144
74C175 2 00
74C192 2 40
74C221 2.75
74C905 3 00
74C906 1.50
74C914 1.95
74C922 7.50
74C923 7.50
74C925 10 50
74C926 10.50
74C927 10 50
.65
2112-2
MK4116
25136
21L02-1
MM5262
MV 53 20
MV533C
P041I0-3
P04110-4
P5101
4200A
B2S25
911 0?A
H001G5-S
MM57100
GIAY38 500-1
MCM657IA
9368
CLOCKS
MM5309
MM5311
MM5312
MM5313
MM5914
MM5315
MM5316
MIM5310
MM5369
MMS841
MM5865
CT7X?
CT7010
CT70I5
MM5375AA/N
MM5375AB/N
P.0. Box 4430 M Santa Clara, CA 95054
will call only: ( 408 > 988-1640
2996 Scott Blvd. ^
Qic/L
f ELECTRONICS
1650
3.75
DS0056CN 3 75
MM53104 2.50
1C SOCKETS
Solder Tin Low Profile
PIN 1 UP PIN 1UP
8 .15 24 36
14 .18 28 .43
16 20 36 .58
COP1802CD
CCP18020
C0P1 861
UARTIEIFO
AV5-1013
AY5-1014
1702 A
N82S23
N82S123
N82S126
N82S129
N82S131
N82S138
N82S137
2708
DM8577
8223
2716
3.85
3 50 4 MHz
6 25 5 MHz
11.50 10 MHz
10 75 18 MHz
19.95 70 MHz
25.00 32 MHz
12 95 32768 Hz
4 25 2.09715? MHz 7 75
4.25 2 4576 MHz ‘ -
4 25 3 2768 MHz
3.90 5.068B MHz
3.90 5.185 MHz
3.90 5.7143 MHz
‘ “ 6.5536 MHz
12.00 1.8432 MHz 4.50 14.31818 ...
15.00 3.5795 MHz ' .20 16 432 MHz
24,50 2.0100 MHz 1 95 22.1184 MHi
RESISTORS Y. watt 5%
5.50 10 per type .03 1000 per type .012
7.50 25 per type C25 350 piece pack
6.95 100 per type .015 5 per type 6.75
KEYBOARDS
4 95 Hex keyboard Si0.9S.fuNy encoded v*/
2.95 PC board, parts and instructs. S29.95
3 50 53 key ASCI' keyboard kit 55 00
3 75 Fully assembled 65.00 Enclosure 14.95
3' 75 LEDS
3 Digit Universal
" Counter Board Kli
4 30 Opeeates 5-18 VoN 0C to 5 MHz
< 30 typ. 125- LEO display 10 50
Voice actuated switch 50
<•?■> Paratronics 106A Logic
4 33 Analyzer Kil S199.00
4 >0 Model 10 Trlgoer
Expander Kit S299.00
el 150 Bus
Red T016
a >5 Green. Orange. Yellow T018
,6 55 -umbo Red
0 an 3'een. Orange, Yellow Jt— k -
- ~ Clipllte LEO Mounting Cl
(specfy red. amber, green
CONTINENTAL SPECIALTIES in stock
.55 .47
MAX-100 6 digit Freq Ctr. $128.95
OK WIRE WRAP TOOLS In stock
Portable Multineles $1 8.00
DIGITAL THERMOMETEH
TRANSFORMERS
12 Volt 300 ma banslormer 125
12.6V CT 600 ma 3 75
■2V 253 ma wall plug 2 95
i2V CT 250 mawaa plug 3S0
24V CT 400 ma 3 95
10V 1.2 amp waa plug 4 85
OISPLAY LEOS
MAM CA .270 2 90
MANS CC 125 39
MAN 72/74 CA'CA .300 1 00
OL704 CC .300 1 25
OL707/OL707R CA 300 1 00
OL727/728 CA/CC .500 1 90
DL747/750 CA/CC .600 1 95
DL750 CC 600 1 95
FND359 CC 357 70
FN0500/S07 CC/CA .500 1 35
FMD503/S10 CC/CA 500 90
... „ FN0800/807 CC/CA .800 2 20
3 dipt Babble 60
2102-1 1.28
2102AL-4 1.60
21F02 1.85
2104A-4 3.95
MICROPROCESSOR
6800 24.50
9080A with data 1150
280 29 95
Keyer 8043
comp. w/spec./so<*
C 44 pin^diK* 2 00 32’-230‘F Disposable probe cover *.2* Di^Vluorescen:
100 pin edge 4 50 accuracy. Comp. assy, in compact case. OG10 Fluorescent
100 pin edge WW 5.25 COMPUTER BOARD KITS
TRANSISTORS Boerd Art
2N2222A 18 « EPROMKit
2N3904 .18 I/O Board tOI
2N3906 18 Extern*'
2N3055
1 75
CRYSTALS
eg Video Interlace boa-d kil
16K EPROM board kit w/o Phi
' 6K Sialic RAW board kit
4.50 North Star Floppy Oisk Kil
4.50 Additional Drive Kit
i digit 14 pel
$134 95 NSN33M 3 digit 8 pin
114.95 NSN69 9 digit display
44 50 7520 Claire* photocells
12 50 T1311 Hex
125 00 S (unction digital DC
MS 74.50 Calendar clock
395.00 .4' green LEO display. 1
$665 00 Pedestal mount. 2.7"
415.00 Beautiful case.
Sinclair 3V2 Digit Multimeter
Batt. oper. 1 mV and . 1 NA resolution. Re-
sistant to 20 meg. 1% accuracy. Small,
portable, completely assem. in case. 1 yr.
guarantee. Best value ever! $59.95
New Cosmac Super “ELF”
RCA CMOS expandable microcomputer
w/HEX keypad input and video output for
graphics. Just turn on and start loading
your program using the resident monitor
on ROM. Pushbutton selection of all four
60 Hz Crystal Time Base
Kit $4.75 Converts digital clocks
from AC line frequency to crystal time
base. Outstanding accuracy. Kit includes:
PC board, MM5369, crystal, resistors,
capacitors and trimmer.
Not a Cheap Clock Kit $14.95
Includes everything except case. 2-PC
boards. 6-. 50" LED Displays. 5314 clock
chip, transformer, all components and
full Instructions. Same clock kit with’ .80"
displays. S21.95
PPM mnHpc 1 PR inHiratnrc nf piirrpnt
vr U IIIUUCo. LlU IIIUILdlUlo Ul LUIICIIl
CPU mode and four CPU states. Single
step op. for program debug. Built in pwr.
supply, 256 Bytes of RAM, audio amp. &
spkr. Detailed assy. man. w/PC board &
all parts fully socketed. Comp. Kit
$106.95 High address display option
Clock Calendar Kit $23.95
CT7015 direct drive chip displays date
and time on .6" LEDS with AM-PM indi-
cator. Alarm/doze feature includes buz-
zer. Complete with all parts, power supply
and instructions, less case.
Digital Temperature Meter Kit
Indoor and outdoor. Automatically
switches back and forth. Beautiful. 50"
LED readouts. Nothing like it available.
Needs no additional parts for complete,
full operation. Will measure -100° to
+200°F, air or liquid. Very accurate.
Complete instructions. $39.95
8.95: Low address display option 9.95;
Custom hardwood cab.; drilled front
panel 19.75 Nicad Battery Backup Kit
w/all parts 4.95 Fully wired and tested in
cabinet 151.70 1802 software xchng.
club: write for info.
4K Elf Expansion Board Kit
with Cassette l/F $79.95
Available on board options: IK super ROM
2.5 MHz Frequency Counter
Kit Complete kit less case S37.50
30 MHz Frequency Counter
Kit Complete kit less case $47.75
Prescaler Kit to 350 MHz SI 9.95
NiCad Batt. Fixer/Charger Kit
Opens shorted cells that won't hold a
charge and then charges them up, all in
one kit w/full parts & instruc. $7.25
monitor $19.95 Parallel I/O port $7.95
RS232 l/F $3.50 TTY 20 ma l/F $1.95
S-100 Memory l/F S5.30
Tiny Basic for ANY 1802 System
Stopwatch Kit $26.95
Full six digit battery operated. 2-5 volts.
3.2768 MHz crystal accuracy. Times to
59 min., 59 sec., 99 1/1 00 sec. Times std.,
split and Taylor. 7205 chip, all compo-
nents minus case. Full instruc.
RCA CosmacVIP Kit 275.00
Video computer with games and graphics.
Kansas City Standard Cassette $10.00
On ROM Monitor $38.00
Super Elf owners take 30%
flrininol Pncmon u PI C 59 LSI
78 1C Update Master Manual
1978 1C Update Master Manual $30.00
Complete 1C data selector 2175 pg. Mas-
ungmai i/osmac CLr kil
All parts and instructs. $89.50
Board only 14.95
Auto Clock Kit $15.95
DC clock with 4-. 50" displays. Uses
MatinriQl MA.1H19 mnHiilp uuifh alarm
ter reference guide. Over 42,000 cross
references. Free update service through
1978. Domestic postage $3.50. Foreign
$6.00. Final 1977 Master closeout $15.00
Video Modulator Kit $9.95
Convert your TV set into a high quality
monitor without affecting normal usage.
Complete kit with full instructions.
IldllUlldl IVIM“ 1 U 1 L. IIIUUUlG Willi dlallll
option. Includes light dimmer, crystal
timebase PC boards. Fully regulated,
comp, instructs. Add S3. 95 for beautiful
dark gray case. Best value anywhere.
TERMS: $5. OOmin. orderU.S. Funds. Calif residents add 6% tax. FREE: Send for your copy of our NEW 1978
BankAmericard and Master Charge accepted. QUEST CATALOG. Include 240 stamp.
L Shipping charges will be added on charge cards. A
/SOLID CAREER OPENING^
WITH HEATH
You know us for our quality Heathkits — and
for our reputation for "doing things right." Our
progress story goes far back, and prospects are
very favorable for the years ahead.
SOFTWARE
DOCUMENTATION WRITER
An effective, proven technical writer needed to
write Heathkit computer systems operations
manuals and applications procedures. A B. Sc.
CS or equivalent is required. Candidates must
be familiar with LSI — 11&PDP-11, software
and applications programming, higher level
languages, and operating systems, principles and
techniques.
Salaries are competitive, with excellent benefits
additional. Our ideal location is a pleasant,
small community on Lake Michigan, 90
minutes from Chicago.
Please send resume in confidence, or phone
collect to Ken Smith, (616) 982-3673.
HEATH
Schlumberger
HEATH COMPANY
49022
Benton Harbor, Michigan
H5
An Equal Opportunity Employer M/E
COMPUTER CLUBS!
Do you need to start a
club library?
Start with the BEST!
Have the secretary of your club
send us your current club roster,
complete with names, addresses
and zip codes and we’ll send you
one of your choice of the follow-
ing books, . . . FREE!
Hobby Computers Are Here
The New Hobby Computers
1C Test Equipment, the new Vol. IV of
the 73 Test Equipment Library
( limit one book per club )
Kilobaud Clubs
Peterborough, NH 03458
120
The All New!
Personal & Small
Business Computer
Plan Expo Plan
To m To
Exhibit ™ Attend
"South"
May 19-21,1978, Exposition Park
Orlando, Florida
For Details Call Or Write:
Felsburg Associates, lnc.(30l)262-0305
P.O.Box 735. Bowie. Md.. 20715
LOOKING FOR THE BEST COMPUTER ON THE MARKET?
WHETHER VOU ARE JUST STARTING OR WANT TO UPGRADE VOUR PRESENT SYSTEM,
WHY NOT HAVE —
THE BEST OF
BOTH WORLDS
6800 COMPUTING POWER
(With the new 6802!)
-AND -
S-100
COMPATIBILITY
YOU CAN
HAVE BOTH
WITH THE
MD690 CENTRAL
PROCESSOR
FROM MDS
AVAILABLE AS SINGLE BOARD OR COMPLETE SYSTEM:
MD690 CENTRAL PROCESSOR BOARD
Single board includes 6802 processor, IK monitor PROM (Standard 6800 compatible), 1152 bytes
of RAM, on-card 2400 baud cassette interface, parallel / serial I/O port — Complete instructions —
Ready to operate — Just plug into any S-100 bus — Interfaces
easily with most keyboards and memory-mapped video cards. KIT $198 ASSEMBLED $269
MDS-2 SYSTEM
Complete system includes case, MD690 processor board, power supply, video / graphics board,
motherboard with two edge connectors (room for 6), and
keyboard with case. KIT $579 ASSEMBLED S798
^UNBEATABLE PRICES***UNEXCELLED PERFORMANCE***
Either board or complete system is supported by extensive MDS documentation and software —
every effort has been made to make this micro easy to understand, easy to use, easy to buy.
‘ [ALL MDS SOFTWARE NOW AVAILABLE BY TELEPHONE DATA LIN Ki|
TO ORDER —
SEND CHECK OR
MONEY ORDER TO:
LAKE ELECTRONICS
Dept. C, Box 45058
Los Angeles, CA 90045
Cal. Res. Add 6% Sales Tax
TH€ COMPUTCR CORNGR
White Plains Mall, Upper Level
200 Hamilton Ave.
White Plains NY 10601
Phone: (914) WH9-DATA
Near Bronx River Parkway &
Cross Westchester Expressway.
Plenty of parking.
"The SI 00 Bus stops at
White Plains" with one of
the largest collections of boards
compatible with the Altair Bus
(also IMSAI) in the greater NY
area*
You've read about the
Sol-20, now come up and
see it. We carry Processor Tech,
Polymorphic, IMSAI, North Star,
TDL, Blast Master and Pickles
and Trout.
GOOD PRICE AND SERVICE
10-6 Mon. -Sat.
Thurs. till 9
C-28
TH€ COMPUTER CORNCR
SURPLUS ELECTRONICS
IBM SELECTRIC^
BASED I/O TERMINAL
(USED) $695.00
• Tape Drives • Cable
• Cassette Drives • Wire
• Power Supplies 12V15A, 12V25A,
5V35A Others, • Displays
• Cabinets • XFMRS • Heat
Sinks • Printers • Components
Many other items
Write for free catalog
WORLDWIDE ELECT. INC.
10 FLAGSTONE DRIVE
HUDSON, N.H. 03051
Phone orders accepted using VISA
or MC. Toll Free 1-800-258-1036
In N.H. 603-885-3705 W16
The Computer Mart
Toronto, Canada
YOUR NUMBER ONE
SOURCE FOR MICROS
AND PERIPHERALS.
PROFESSIONAL SERVICE
AND GUIDANCE.
CATALOG $2.00
1543 Bay view Avenue,
Toronto, Ontario M4G3B5
[416] 484-9708
121
WORLD’S
RECORD^ HOLDER
ATE* assembler/text editor
by Soft Corp is the most power-
ful text-oriented software ever
created for 8080- based per-
sonal computers.
ATE* is an interactive text
editor, a fully symbolic assem-
bler, a system monitor, a fully
programmable text-oriented
language, and a cassette
operating system — all in one
4K package.
In short, ATE* contains
everything you need in soft-
ware to create computer files
for your personal or business
records of names, telephone
numbers, dates, addresses and
the other textual facts of life.
Then, using ATE's* powerful
programming language, you
can build your own programs
to put your files to work in
exciting, highly personalized
applications.
ATE* contains so many
features to make text-oriented
operations simple, logical and
efficient that you'll have to see
TEXT EDITOR
ASSEMBLER
PROGRAMMING
TAPE HANDLING
it to believe it. Editing by
character, string or line. Inter-
active and programmable
editing. Printing, editing and
assembling object code in your
choice of octal, hex or decimal.
And so much more we can't
even begin to list ATE's* full
array of features.
Ask to see ATE* demon-
strated at your nearest computer
store. Or ask your local com-
puter store to order it for you.
ATE* on Kansas City Standard
300 baud cassette with Users'
Manual, $25.t
If unavailable locally, order
direct from Thinker Toys™,
1201 10th St., Berkeley, CA
94710. Call (415) 527-7548 (10- 4
PST). Mail and phone orders
add $3; Cal. res. add tax.
A product of Soft Corp for
Thinker Toys ■
1201 10th Street Berkeley, CA 94710
T28
♦ATE is a tm of Soft Corp.
tDeutsche iibersetzung von computershop Gmbh, D-7800, Freiburg im Breisgau, Adelhauserstrasse 29, W. Germany
LOGOS I
8K STATIC MEMORY
BOARD
T ~r=^COMPOTER
^ /PRODUC TS
uiiiiiji
SPECIFICATIONS:
System Compatibility: S-100 buss compatible.
Altair/lmsai compatible.
Memory Protect: Hardware memory protect circuit features protection of board
as one 8K byte block, two 4K blocks, four 2K blocks, eight 1 K
blocks sixteen 512 byte blocks, or thirty-two 256 byte blocks,
or disable entirely all options are dipswitch selectable.
Addressing: Addressing on any 1 K boundary.
Wait States: No Wait States
Buffering:
Speed :
Support Chips:
Truly buffered. Address lines and Data-in and Data-out fully
Buffered. Key feature as most 8K memory boards are not fully
buffered.
Logos I — 450 ns access. Allows board to run at full
speed with no wait states.
Logos IZ — 250 ns access. Allows board to run at full
4MHz Z-80 speed.
Low power Schottky (where applicable)
PC Board
Hi-grade gloss epoxy with plated thru holes, gold-plated edge
connector contacts, solder-mask, with silk screen.
Power:
+8v, 1.5 Amps typical
Battery Back-up:
Circuitry on board
ORDER INFORMATION
Logos I 8K Kit $125.95 Buy (4) Units reduce to $117.00 Imsai Compatible Wirewrap Connector $4.75
Logos I 8K Assembled & Tested $179.95 Imsai Compatible Solder-tail Connector $4.75
Logos I 8K-Z (250ns.) Kit $149.95 Altair Compatible Wirewrap Connectors $5.95
Logos I 8K-Z Assembled & Tested $199.95 Altair Compatible Solder-tail Connectors $5.95
ADVANCED COMPUTER PRODUCTS INC.
Irvine, CA 92713 • P. O. Box 17329 • (714) 558-8813 • TELEX/TWX: 910-595-1565
A39
ENTERPRISES
Incorporated
P.O. Box 254 King of Prussia PA 19406
(215) 279-7968
HEAVY DUTY IBM 735 I/O SELECTRIC BASED TERMINAL
MECHANICS COMPLETELY REFURBISHED—
ALL NEW MICROPROCESSOR-BASED ELECTRONICS-
SWITCH SELECTABLE FOR EITHER ASCII OR EBCDIC OPERATION
PRICE: $895
! WHAT YOU SHOULD KNOW BEFORE BUYING ANY STANDARD IBM SELECTRIC TERMINAL
• Carriage Return time is about 5 times longer than a standard terminal; therefore, you need to transmit up to 12 null or rubout
characters after the standard CR/LF characters to allow enough time for the carriage return. This may require you to rewrite your
computer’s software. There are other characters which have similar problems such as Index, Tab, Backspace and Shift.
• The mechanics of the IBM Selectric limit the printing speed to a maximum of 14.9 characters per second, therefore it cannot
run at 150 baud (15 characters/sec.)
• The standard baud rate for a Selectric is 134.5 and therefore cannot interface with a system having only the standard baud
rates such as 110 or 150 without modifying or completely replacing the terminal’s electronics.
• Some of the IBM Selectric terminals use a unique character ball and are not interchangeable with the standard typewriter ball.
The balls for these are more expensive, harder to find, and do not have the font selection.
• The IBM Selectric’s printer and keyboard are mechanically linked together and therefore, without sophisticated electronics, it
cannot interface with a full-duplex system.
• The Selectric produces only 10 standard control codes versus 34 on a standard ASCII terminal.
• There are several IBM Selectric terminals around and not all have the heavy duty Selectric mechanism.
WHAT YOU SHOULD KNOW ABOUT THE IBM I/O SELECTRIC TERMINAL THAT WE ARE OFFERING
• TWO operating systems (switch selectable)
A. As a standard IBM terminal using EBCDIC Code at 134.5 Baud. So that it can be used with IBM equipment.
B. As a Full 7-bit ASCII terminal at 110 Baud. With the following features;
1. The terminal operates at 10 cps, but prints at 14.9 cps and has a 150 character buffer to compensate for the long carriage
return time. Therefore there is no requirement to rewrite your computer’s software.
2. It generates all 34 of the standard ASCII control codes.
3. Full or Half-duplex operation.
4. Generation of parity.
5. Two modes of operation:
a. TTY Mode: Transmits only upper-case codes (for alpha characters only) like a standard TTY even if the shift key is not
depressed.
b. Typewriter Mode: Transmits both upper and lower-case codes, dependent upon the shift key being depressed or not.
• Has both RS-232 and 20 ma. Current Loop interfaces.
• Remote/Local switch, so it can be used as a typewriter or a terminal.
• Uses the standard IBM Selectric character ball.
• Has a 15” carriage for up to 132 characters per line.
• Platen feed.
=Z I ' ~ ALSO AVAILABLE ' "" "
Custom Power Supply designed for the KIM-1, providing 5vdc @ 1.2 amps & 12 vdc @ .1 amps. Price: $40.00, plus $1.50 shipping
& handling. Commercial duty— Full 2 year warranty.
COMING SOON
A PROM blower for 2708s and a PROM card for 2708s, 2758s, or 2716s, and Mini-2 Slot Mother Board and 8K RAM Board— all
designed for the 6502 based KIM-1.
• ALLOW 6 TO 8 WEEKS FOR DELIVERY • PRICE INCLUDES FULL DOCUMENTATION • 30 DAY WARRANTY— PARTS AND LABOR
Terminals only, select: □ Airfreight DSurface terminals shipped freight collect— fob Phoenix az
Enclosed: Dcheck DM.O. Charge DVISA □ Master Charge
Card #. Interbank #
Expiration date:.
Signature:.
# OF TERMINALS @ $895
# OF POWER SUPPLIES @ $41.50
PA residents must add 6% sales tax
Total amount of this order $
NAME:
$_
$_
$_
$_
ADDRESS:
CITY:
.STATE:.
.ZIP:
PHONE: ( ).
Visa (BankAmericard) & Master Charge Accepted.
124
S.D. COMPUTER PRODUCTS
AN EMPIRE IND CO
S2
P.O. BOX 28810K
DALLAS, TEXAS 75228
EXPANDABLE EPROM BOARD
16K OR 32K EPROM $49.95 W/OUT EPROM
Allows You to Use Either 2708's For 16K of Eprom or
2716 s For 32K of Eprom.
KIT FEATURES:
1. All Address Lines & Data Buffered
2. Quality Plated through P.C. Board Including
Solder Mask and Silk Screen
3. Selectable Unit States
4. On Board Regulation Provided
5. All Sockets Provided W/Board
WE CAN SUPPLY 450As 2708's AT $11.95
WHEN PURCHASED WITH BOARD.
4K LOW POWER RAM KIT
The Whole Works -$79.95
Full Buffered - on board regulated - reduced
power consumption utilizing low power 21L02-1
500ns RAMS - Sockets provided for all IC's.
Quality plated through PC board.
•Add SI 0.00 for
2SOns RAM operation.
EXPANDORAM
THE ULTIMATE RAM BOARD
32K FOR $475.00
32K MODEL
8K $151.00
16K 259.00
24K 367.00
32K 475.00
THE 32K VERSION USES THE MOS-
64K MODEL
16K $281.00
32K 519.00
48K 757.00
64K 995.00
TEH MK4115 RAM AND HAS 8K rn maximum nF ' Ask ,t,v n«a ion
5?!uzi 1S E DIpTw.K 0 H T E E S CT,0 P N C 4 S?Xy ?N STEP? OF EITHER 8K BOUNOARIES AND PROTECTiqN A
«K R nPFRATmN WITH S0CKETS F0R ly purchasing* more° N ra B m chTps
3ZK OPERATION FROM SO COMPUTER PRODUCTS.
UTILIZES DIP SWITCHES.
BOARD COMES WITH SOCKETS FOR
64K OPERATION
LOOK AT THE FEATURES WE NAVE BUILT INTO THE EXPANDORAM!
• NO WAIT STATES REQUIRED
• NO CYCLE STEALING NEEDED
• ON BOARD REGULATION
• CONTROL. DATA & ADDRESS INPUTS
UTILIZE LOW POWER SCHOTTKY
DEVICES
• DESIGNED TO WORK WITH Z-80,
8080. 8085, CPU's
ADD $50 00 TO ABOVE PRICES FOR FULLY ASSEMBLED AND TESTED BOARDS
» MEMORY ACCESS TIME IS 375 ns
• MEMORY CYCLE TIME IS 500 ns
» POWER REQUIREMENTS ARE:
8 VDC 400 MA OC
18 VDC 400 MA DC
—18 VDC 30 MA DC
► ON BOARD INVISIBLE REFRESH
Low Cost
Cassette
Interface Kit
$19.95
Fea.jres: Play and record K.C. Standard 2400/1200 Ha
tapes, 300 Baud, TTL 1/0 Compatible, Phase Lock Loop,
Both 22 Pin Connector and 8 Pin Molex Connector.
Comes partially assembled Oscillator end phase lock
.oop pre tuned to K.C. Standard. Selector switch sends
cassette data or auxiliary input data to microprocessor.
LED indicates logic 1 level.
8K LOW POWER RAM
$159.95
FULLY ASSEMBLED AND TESTED. NOT A KIT.
Imsai — Altair — S-100 Buss compatible, uses low power
static 21L02-50Qns fully buffered on board regulated,
quality plated through PC board, including solder mask. S
pos. dip switches for address select
•Add $30.00 for
250ns RAM operation
Z-80 CPU BOARD KIT
Complete Kit $139.
Hjl »|
CHECK THE ADVANCED FEATURES OF OUR Z-80
CPU BOARD: Expanded set of 158 instructions,
8080A software capability,
operation from a single
5VDC power supply: always
stops on an Ml state, true
sync generated on card (a
real plus feature!), dynamic
refresh and NMi available,
either 2MHZ or 4MHZ op-
eration, quality double sided plated through PC
board; parts plus sockets provided for all IC’s.
*Add $10. extra for Z-80A chip which allows
4MHZ operation.
NEW FROM S.D.
“VERSAFLOPPY”™ KIT
THE VERSATILE FLOPPY DISK CONTROLLER
ONLY $149.00
FEATURES: IBM 3740 Soft Sectored Compatible. S-100 BUS
Compatible for Z-80 or 8080. Controls up to 4 Drives (single
or double sided). Directly controls the following drives:
1. Shugart SA400/450 Mini Floppy
2. Shugart SA800/8S0 Standard Floppy.
3. PERSCI 70 and 277.
4. MFE 700/750.
5. CDC 9404/9406.
34 Pin Connector for Mini Floppy. 50 Pin Connector for Stand-
ard Floppy. Operates with modified CP/M operating system
and C-Basic Comptler. The new "Versafloppy” from S.D.
Computer Products provides complete control for many of
the available Floppy Disk Drives, Both Mini and Full Size.
FD1771B-1 Single Density Controller Chip, Listings for Con-
trol Software are included in price.
FD 1771B-1 CHIP ALONE $39.95
S. D. SALES NEW
EXPANDABLE EPROM BOARD
16K OR 32K EPROM $49.95 W/OUT EPROM
Allows You to Use Either 2708’s } For 16K of Eprom or
2716’s For 32K of Eprom.
KIT FEATURES:
1. All Address Lines & Data Buffered
2. Quality Plated through P.C. Board Including
Solder Mask and Silk Screen
3. Selectable Unit States
4. On Board Regulation Provided
5. All Sockets Provided 3 W/Board
WE CAN SUPPLY 450As 2708’s AT $11.95
WHEN PURCHASED WITH BOARD.
INTRODUCING THE SBC-100
(The Z-80 Based, S-100 Single Board Computer)
$349.00
FEATURES:
• No Front Panel Needed
• Z-80 CPU (2 or 4 MHZ)
• IK RAM
• 4 ROM/PROM Sockets for 4K/8K of Memory
• SYNCHRONOUS/ASYNCHRONOUS Serial I/O with
RS-232 and Current Loop Interface and Software
• Programmable Baud Rate
• Parallel Input Port
• Parallel Output Port
• 4 Channel Timer/Counter
• 4 Vectored Interrupts
O. E. M. SPECIAL
ASK ABOUT SPECIAL O E M. DISCOUNTS ON THE S.D.
SOC-lOO — SINGLE BOARD COMPUTER
VERSAFLOPPY^ — FLOPPY DISK CONTROLLER
EXPANDORAM — 32K RAM
•COMPATIBLE SET
$349 00
$149.00
$475.00
EACH KIT IF PURCHASED SEPARATELY TOTAL $973.00
ORDER ALL 3 KITS TOGETHER FOR
$899.00
Thi* Powerful Threesome Operates Together to Form A Comolete Computer for Your System.
Z-80
Programming Manual
IN DEPTH DETAIL OF
THE Z-80 CPU
MICROCOMPUTER
S. D. SALES SPECIAL
$9.95
RAMS
2lLu2 - 500NS
21L02 - 250NS
2114 — 4K
1101 A —256
1103 — IK
MK 4115-8K
74S 200 256
8/11 50
8/15 95
14 95
8/$4 00
35
16 45
3.95
CPU’s
Z — 80 includes manual
Z — 80A includes manual
8080A CPU 8 BIT
8008 CPU 8 BIT
29 95
34 95
11 95
6 95
PROMS
1702A - IK - 1.5 us . .3.95 or 10/35.
2708 - 8K - 450ns 14.95
5204 • 4K 7.95
82S129 — IK 2.50
2708U 8K signetics 650ns 9.95
COUNTER CHIPS
MK50397 6 Digit elapsed timer
MK50250 Alarm clock . .
MK50380 Alarm chip .
MK50395 6 digit up/dn. count
MK5002 4 digit counter
MK5021-Cal. chip sq. root
8.95
4.99
.2.95
12.95
8.95
.2.50
S.D. NOW HAS SOFTWARE FOR IT’S CUSTOMERS
CP/M 1 DISK OPERATING SYSTEM
$99.95
CP/M is a powerful disk operating system which has become an industry standard. It is compatible with several
disk based FORTRAN and BASICS This package includes a CP/M diskette (mini or full size) adapted for S.D.'s
SBC lOO/VERSAFLOPPY EXPANDORAM board set. Complete documentation is included. ( u CP/M is a reg-
istered trademark of Digital Research Corp., Pacific Grove. CA.
Z-80' DISK BASED ASSEMBLER $69.95
Runs on ANY CP/M based disk system. Assembles
the official Zilog-Mostek Mnemonics. Contains ex
tensive set of pseudo-ops. Available on mini or
full size diskette.
VERSAFLOPPY ' 1 CONTROL FIRMWARE . . . $24.95
Provides control tor VERSAFLOPPY and boots up
CP/M. This runs on Z-80. 8080 or 8085 based
computers. Available in 2708 or 2758 prom.
$49.95
SD MONITOR
Powerful monitor for SBC-lOO single board com-
puters. Includes all VERSAFLOPPY control firm-
ware. Comes in 2716 prom. Available in 4-6 weeks.
VERSAFLOPPY DIAGNOSTIC PROGRAM . . . $24.95
Provides routines which are helpful in checking out
a disk based system Available in 2708 or 2758
prom.
★
SUPER FLOPPY SPECIAL
S. D. SALES' VERSAFLOPPY S-100 CONTROLLER BOARD PLUS
SHUGART SA 400 FLOPPY DISK DRIVE INCLUDING CABLE FOR ONLY
$449.00
★
MICROPROCESSOR
CHIPS
8212 • 1/0 port
. .3.50
8214 — P.I.C.
.12.95
8216 — Non Invert But ...
. .4.95
8224 — Clock Gen
. .4.95
8226 — Invert Bus
. .3.95
PIO for Z— 80 .
.14.95
CTC for Z— 80 .
.14.95
8228 Sys. Controller
8.20
8251 Prog. comm, interfect.
.10.95
8255 prog. prep.
interface . .
13.50
8820 Duel Line Recr
.1.75
8830 Dual Line
Dr
..1.75
2513 Cher. Gen.
. .7.50
8838 Quad Bus.
Recvr. . . .
.2.00
74LS138N — 1/8 decoder
. . ..99
8T97-Hex Tri-State Buffer .
.1.25
1488/1489 RS232
.1.50
TR 1602B Uart .
.3.95
TR 18S3 Uart .
8.50
FD 1771B-1 ...
.39.95
CMOS
4001 19
4029
99
4002 19
4042
69
4011 19
4047
1.50
4013 32
4049
35
4016 32
4069
23
4017 95
4071
19
4020 97
4076
97
4022 97
14518
1.10
4024 75 .
14528
85
4027 39
14529
85
CALL IN YOUR BANKAMERICARD
(VISA) OR MASTER CHARGE OR-
DER IN ON OUR CONTINENTAL
TOLL FREE WATTS LINE:
1 — 800 — 527—3460
Texas Residents Call Collect:
214 / 271-0022
DEALER INQUIRIES INVITEDI
(All prices subject to change
without prior notice.)
NO COD s. TEXAS RESIDENTS ADD
5% SALES TAX. ADD 5% OF ORDER
FOR POSTAGE & HANDLING . OR
DERS UNDER $10. ADD 75c HAND-
LING. FOREIGN ORDERS - U. S.
FUNDS ONLY!
125
CaUFoRnIa iNduSTRiAl
Post Office Box 3097 K • Torrance, California 90503
Diablo terminal
In February, when this ad-
vertisement was submitted
to the publisher, we were
negotiating for the pur-
chase of several hundred
used Diablo Terminals.
If we are successful in
acquiring these units, they
will be available in late
April for only $850.00. FOB
Los Angeles.
All terminals were removed
from service in operating
condition.
Edge
Connectors
100 PIN
IMSAI/ALTAIR
S-100 • GOLD PLATED • .125" CENTERS
Altair .140 row, soldertail $5.98 3/S16.50
Imsai .250 row, soldertail $4.98 3/$13.00
3 Level Wire Wrap .250 row. . . $4.98 3/$ 13.00
SPECIALS
W/W same as above without ears$3.50 3/$10
72 (dual 36) W/W .156" ce nters. . . $2 .50 3/$6
Certified Digital
Electronic Entertainment Center
Tennis-Handball
Hockey-Smash
Color *24.88
Action-packed color entertainment tor
the whole family. Adjustable skill level
controls allow players of all ages to com-
pete in tennis, hockey and handball.
This four game entertainment center
turns your television into a video play-
ground.
On screen scoring, live action sound and
true component color makes th'is video
center an excellent buy at only $24.88.
Complete with antenna box and AC adapter.
HEXADECIMAL KEYBOARD
Maxi-Switch hexadecimal keyboards are designed for $34.95
microcomputer systems that require 4-blt output
in standard hex code.
Each assembly consists of 16 hermeti-
cally sealed reed switches and TTL "one J
shot” debounce circuitry.
Reliable low friction acetal resin
plungers are credited for the smooth
operation and long life of this premium
keyboard.
Requires single + 5 volt supply.
‘24.88
ivo ^UNIVAC
KEYBOARD
The famous Sperry Univac 1710 Hollerith keyboard assembly
is now available from California Industrial tor only $24 SB
The ideal computer input device for accountants and
mathematicians The numeric keys are placed on the lower
three rows to resemble a ten key adding machine This
TELETYPE MODEL 43
New from Teletype, the Model
43 is capable of printing 132 ASCII
characters per line. Send and receive
data at 10 or 30 char, per second. Key-
board generates all 128 ASCII code combina-
tions. RS-232 interface, same as the popular
Model 33. Data sheet sent upon request. Manufac-
turer suggested price $1377.00.
IMMEDIATE DELIVERY *1219
TTL model with NOVATION brand
Acoustic Modem. *1419
tst
REGULATED
POWER SUPPLY
Delivers 5 volts at 8 Amperes
along with three other regu-
lated outputs.
This used supply is sold “as
is,” but we still feel that this
is the best buy you’ll ever see
in a regulated power supply.
75 lbs., Schematics included.
Used *49.50
Shipped Freight Collect I
Quiet Buss
S-100 MOTHER
The Quiet Buss from California Industrial is
quality engineered. No short cuts have been taken
to produce this mother board. Active termination
circuitry prevents noise and irosstalk. Manufac-
tured from extra heavy FR-4 epoxy glass. Features
2 ounce double thickness copper traces.
^29.95
Lii
-
mnnuflL graphite
display GEnERATDR |
Modern technology has pioneered the development ot this I
unique character printer. Our Manual Graphite Display I
Generator has the capability ol producing the full upper and I
lower case ASCII set Self-contained cursor assembly allows I
the operator to eliminate erroneously entered information P
Each unit is manufactured to strict tolerances as prescribed |
by standards set (orth by California Industrial One tree with I
every order
SCOtCH CASSETTES H*
Please specify
IBM 3740 series
or 32 sector,
also
MINIDISKS
Won't drop a BIT!
CALIFORNIA
INDUSTRIAL
is an
Authorized
Dealer of
Scotch Brand
Data Products
■ WW' ■-
L *1000
This joystick feature four 100K potentio-
meters. mat vary resistance proportional to
the angle of the stick. Perfect (or television
games, quad stereo and radio controlled
aircraft
9?5
5&l2voU
REGULATORS
10 100
*l?-9ea. .97 .88
Negative |
Positive
UV Erasable
T0-3 1
LM320KO
LM340KO
MEMORY
TO -220 1
1 LM 320TO |
LM340K )
S-100 PROTOTYPE BOARD |
*19.98
GPIOO-Moximum design ver-
sotlllty along with standard
address decoding ond buf-
fering for SlOO systems.
Room for 32 uncommitted 16
pin IC's, 5 bus buffer & de-
coding chips. 1 DIP address
select switch, o 5 volt regu-
lator and more.
WWlOO-vv re wrop bread-
board, similor to the GP100.
Allows wire wrap ol ail sizes
of sockets In any sizes ot
sockets in any combination.
An extra regulator position
for multiple voltage applica-
liens
Tbuymbwheel
switch
^ |Ten position
Miniature /
Switches
your choice
q Q 10 50 100 lk
5.9 o $. 88 . 81 .73 .66 I
SPOT Miniature Toggles
7101 C&K ON N0NE 0N
7107 jbt ON-OFF(mnt.ON) I
7108 CK ON-(moment.ON)
Rocker JBT DPDT
Rotary 3P-4-Pos.
Rotary 3P-6-Pos.
Push B (N.0.) $.39ea. 3/S1
MINISCOPE $ 289.00 NLS MS-15 MINISCOPE $ 289.00 NLS
• Includes etciwd circuits and (nsirudions lot option ol
Price:
$ 29.50
8800V
Universal Microcomputer/processor
plugboard, use with S-100 bus Com-
plete with heat sink & hardware b 3" x
10 x 1/16
1-4 5-9 10-24
$19.95 $17.95 $15.96
8801-1
Same as 8800V except plain: less power
buses & heat sink.
1-4 5-9 10-24
3682 9.6” x 4.5”
$10.97
3682-2 6.5” x 4.5”
$9.81
Hi-Density Dual-In-Line
Plugboard for Wire Wrap
with Power & Grd. Bus
Epoxy Glass 1/16” 44
pin con. spaced .156
1/16
Gen. Purpose D.I.P.
Boards with Bus Pattern
for Solder or Wire Wrap.
Epoxy Glass 1/16” 44
pin con, spaced .156
3662 6.5” x 4.5”
$7.65
3662-2 9.6” x 4.5”
$11.45
P pattern plugboards for
IC's Epoxy Glass 1/16”
44 pin con. spaced .156
R644 P C recepticle 22/44 cont
156 ctrs pierced solder eyelets tails
gold 53.45
R644-3 P C recepticle 22/44 cont.
156 ctrs Wire Wrap tails gold $4.49
S-100
Bus P.C. Edge Connectors
R681 DIP solder tails on 140' spaced
rows tor ALT AIR mother boards Fits
.042" dia holes, gold $7.35
R681-1 .025" sq., 3 wrap lengths
posts (.62 long) on 250 spaced rows.
gold $5.00
R681-2 025" sq.. 1 wrap length posts
( 22 long) on .250 spaced rows for wrap-
ping or DIP solder tor IMSAI mother
boards, gold $5.85
R681-3 pierced solder eyelet tails, gold
$7.35
BOARD
Phenolic
.042 dia holes on
0.1 spacing for IC’s
PART NO.
SIZE
PRICE
64P44-062XXP
4.5 x 6.5"
$ 1.49
169P44-062XXP
4.5 x 17"
$ 3.51
Epoxy Glass
PART NO.
SIZE
PRICE
64P44-062
4.5 x 6.5”
$ 1.70
84P44-062
4.5 x 8.5"
$ 2.10
169P44-062
4.5 x 17"
$ 4.30
169P84-062
8.5 x 17"
$ 6.39^
Wraps insulated wire on 025" square posts
FOUR TIMES FASTER
*9
with two
100' spools
of 28 oa.
$24.50
pm®
P160-4T
Includes
charger, wire
$75.00
NO PRE STRIPPING*
NO PRE-CUTTING*
SPOOL-FED WIRE*
*Tl»e spooled wire passes through the loo* past a s itting edge
next to the wrap post A narrow longitudinal cut is made in the
insulation where >t presses the square post
tn« tu rn coop*' is ind«rud s* me slurp sooe
(7 turns- ?8 COMMi)
SLIT-NWRAP
WIRE
NO. 28 GAGE INSULATED
WIRE, 100' SPOOLS
W28 2 A Ph9 3 Green W?8?CPkg 3. Oar
W28-2-8 Peg 3 Red W?B ? 0-PVg 3. Blue
2708
8K 450 ns
EPROM
FACTORY PRIME
1-9 $12.00
10-24 $11.50
25 + Call for Price
r
14 & 16 PIN
3 LEVEL
WIRE WRAP
SOCKETS
14-T3 100 for
$30.00
16-T3 100 for
$30.00
50 of ea. for $32.00
ACE • All- tor fast, solderless.
Circuit plug-in circuit building
Evaluator and testing
iy components with leads to .032" dia. Inter-
olid wire up to 200a. Assembled models tool
LIQUID CRYSTAL DIGITAL
CLOCK-CALENDAR
• For Auto, Home, Office
• Small in size (2x216x14)
• Push button for seconds release for date.
• Clocks mount anywhere with either 3M double-
sided tape or VELCRO, included
• 2 MODELS AVAILABLE:
LCD-101, portable model runs on self-contained
batteries lor better than a year
LCD-102, runs on 12 Volt system and is back-
lighted.
• LCD-101 or LCD-102 CQA QC
your choice #0‘4.UJea
• Clear desk stand lor $2.00
fr
=| 3 LEVEL
Jt GOLD WIRE
WRAP SOCKETS
1
1 124
25-49
50-99
8
pin
.47
.42
.37
10 pin
.45
.41
.37
14 pin
.39
.38
37
16 pin
.43
.42
.40
18 pin
.75
.68
.62
22 pin
1.00
.97
.95
24 pin
1.00
.94
.75
28 pin
1.09
.98
.84
36 pin
1.59
1.45
1.30
40 pin
1.49
1.39
1.29
4
LED ALARM CLOCK
12 hr. LED Alarm Clock uses 3'/?
digit 8‘' LED Display with AM/PM
indicators and colons. Direct drive.
PIN to PIN interlace withsi998A
I.C. Just add switches, AC Supply.
Alarm Display and 1C. only
$7.95 or 2/$15.00
PRIORITY
Price Breakthrough!
$ 1 7 95
MA1003
CAR CLOCK
Bright Green Fluorescent Display Crys-
tal Time Base Assembled, just add
switches and 12 VDC.
SPECIAL
^14CS2100 for *14 00
- 16CS2100 for *16 00
II 14 pin CS2 10 for *2"
m Tm 16 pin CS2 8 for ‘2*'
These low cost DIP sockets will accept
both standard width plugs and chips.
For use with chips, the sockets offer a iow
profile height of only .125" above the board.
These sockets are end stackable.
ELECTRONICS
10031 W Woodley A/e 0 Sepulveda CAo 91343
Terms: VISA. MC. BAC, check, jVloney Order. C.O.D., U.S. Funds Only. CA residents add 6% sales tax Mini-
mum order $10.00. Orders less than $75.00 include 10% shipping and handling; excess refunded. Just m case
please include your phone no. "Sorry, no over the counter soles* 1 Good thru June 15. 1978
phone orders welcome (213) 893-8202
24 PIN DIP PLUGS
WITH COVERS
OEM and Institutional inquiries invited.
BdOOSINIIAI 9L-SIAI SIN
NLS MS-15 MINISCOPE $ 289.00 NLS MS-15 MINISCOPE $ 289 .
U * ••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••++*+°°**°°???**«« r
#••••*
#•••«•
#•••..
#••••-
##•••*
#•••«*
#•••*•
••••*-
##••*•
•••«**
#•••*>
#••••■
#••••*
••••*'
##•••
#•••.-
#•••-
#•••*•
#•••»'
#•••*
#•••-
#•••*'
#•••*
••••«>
••••**
••••*’
#••*•-
#•••-
#•••*>
••••»-
••••*-
•••*"
#••••’
••••«’
••••*'
##•••-
#•••**
#••••*
••••**
#•••-
••••«*
#•••**
#•••*'
#•••«>
#•••-
••••*'
#•••-
#•••«'
••••**
#•••-
• ••*•'
#•••**
#•••-
• •••**
#•••-
#•••-
••••**
§#••••
•••#*>
#•••*•
• •••«*
#•••**
#•••’*
#•«••*
• •••»'
• •••*-
§•••-
#•••*•
##•••*
#••••*
#••••*
• •••«'
• •••**
#•••-
##•••*
#•••*'
#•••“
##•••♦
#•••-
••••*«
#••••*
••
Tlew!
KIT FEATURES:
16K E-PROM CARD
IMAGINE HAVING 16K OF SOFTWARE ON UNE AT ALL TIMEI
S-100 Ilmsai/Altair] Buss Compathle!
1. Double sided PC board with solder
mask and silk screen and gold plated
contact fingers.
2. Selectable wait states.
3. All address lines & data lines buf-
fered!
4. All sockets included.
5. On card regulators.
KIT INCLUDES ALL PARTS AND
SOCKETS (except 2708's). Add $25. for
assembled and tested.
DEALER INQUIRIES INVITED!
Our 2708's (450NS) are $12.95
when purchased with above kit.
Swtw
BK LOW POWER RAM KIT-S149.00
S-100 (Imsai/Altair) Buss Compatible!
KIT FEATURES:
1. Doubled sided PC Board with solder
mask and silk screen layout. Gold
plated contact fingers.
2. All sockets included.
3. Fully buffered on all address and
data lines.
4. Phantom is jumper selectable to
pin 67.
5. FOUR 7805 regulators are provided
on card.
USES 211.02 RAM'S!
Fully Assembled & Burned In
$179.00
Blank PC Board w/ Documentation
$29.95
Low Profile Socket Set 13.50
Support IC's (TTL & Regulators)
$9.75
Bypass CAP'S (Disc & Tantalums)
$4.50
MOTOROLA QUAD OP - AMP
MC 3401. PIN FOR PIN SUB.
FOR POPULAR LM 3900.
3 FOR $1
RECTIFIER SPECIAL
1.5 AMP 100PIV
EPOXY CASE AXIAL LEADS
15 FOR $1
MOTOROLA 7805R VOLTAGE REGULATOR
Same as standard 7805 except 750 MA output.
TO— 220. 5VDC output.
44c each or 10 for $3.95
FULL WAVE BRIDGE
4 AMP. 200 PIV.
69C EA. 10 FOR $5.75
NOT ASSOCIATED WITH DIGITAL
RESEARCH OF CALIFORNIA, THE
SUPPLIERS OF CPM SOFTWARE.
450 NS! 2708 EPROMS 450NS!
Now full speed! Prime new units from a major U.S. Mfg. 450 N.S.
Access time. 1 K x 8. Equiv. to 4-1 702 A's in one package.
Special Offer: $12.95 each when
$15.75 63. purchased with our 16K EPROM kit!
Z— 80 PROGRAMMING MANUAL
By Mostek, The major Z-80 second source. The most detailed
explanation ever on the working of the Z-80 CPU CHIPS. At
least one full page on each of the 158 Z-80 instructions. A MUST
reference manual for any user of the Z-80. 300 pages. Just off the
press! A D.R.C. exclusive! $12.95
POWER RECTIFIER #2 Motorola
Stud Mount. IN1 187.
35 AMPS. 300 PIV.
Military Quality! $ 1.19 e a. or 4/S3.50
4K STATIC RAM'S
2114. The new industry
standard. Arranged as IK
x4. Equivalent to 4-21
L02's in 1 package! 18
pin DIP. 2 Chios give 1Kx8.
2/$24. 8/$85.
741COP AMPS
Mini Dip. Prime new
units. Has computer
Mfg's house number.
12/$2. 100/$15.
Jumbo Red Leds
New by G.E. Like
MV5024. Number
SSL-22.
6/$1. 25/$3.75
OPCOA LED READOUT
SLA-1. Common Anode.
.33 inch character size.
The original high efficiency
LED display. 75c ea.
or 4 for $2.50
Disc Capacitors
.1 MFD 16V. P.C.
leads. Most popular
value. By Sprague.
20/$ 1.00
12VDC Relay by
CDE. SPDT. Coil
is 280 ohm. Small
Size. $1.29
CMOS SPECIAL!
CD 4001 - 5 for $1 .
CD 4011 — 5 for $1.
CD 4013 -3 for $1.
CD4040 — $1. each
CD4042 - 2 for $1 .
CD 4049 — 3 for $1.
3 AMP RECTIFIER
1N4721. Axial Lead. 200 PIV.
HEAVY DUTY!
House numbered
4 for
Metal
Case
$ 1.00
EXPERIMENTER S CRYSTAL
262.144KHZ. This frequency is
2 to the 18th power. Easily
divided down to any power of 2,
and even to 1HZ. New by CTS-
Knight. A $5. value!
$1.25 each
HEAVY DUTY!
Full Wave Bridge
25AMP 50PIV
$1.25
GE Ni-Cad Battery Pack
3 Cell pack, gives 4 volts
at 900MAH. Brand new,
factory fresh. Each cell is
2/3 "C" size. $4.50 /pack.
Buy 3 packs (12 volts)
for $10.95. Limited stock!
Tantalum Capacitors
1 MFD. .35V. By
Kemet. Axial Lead.
Best Value! 10/$1.
GE 10 AMP Triac
SC146D. House no.
To-220 case. Rated
10 amps 400PIV.
75cea. 3/$2.
Digital Research Corporation
(OF TEXAS) ■
P. O. BOX 401247 • GARLAND, TEXAS 75040 • (214) 271-2461
2N3904-House No.
TO— 92. NPN.
VCEO-45.
HFE 100 to 300
10 for $1.00
LS SERIES TTL
74LS00 — 33c 74LS74 — 49c
74LS02 — 35c 74LS90 — 69c
74LS04 — 35c 74LS138— 89c
74LS08 — 35c 74LS154— 1.49
74LS10 — 33c 74LS175— 1.10
74LS20 — 33c 74LS367— 75c
74LS73 — 49c 74LS368— 85c
D20
TERMS: Orders under $15. add 75c. No COD's. We
accept VISA, MasterCharge and American Express
Cards. Money Back Guarantee on all items! Texas
Residents add 5% Sales Tax. WE PAY POSTAGE!
••••#
»*•••
»••••
•••if
•••I#
*••••#
*••••#
••••#§
*••••#
*••••#
••••#
*••••#
••••ft
••••ft
••••ft
••••ft
••••ft
*«••••
•••ft
••••ft
••••ft
••••ft
•••ft
•••ft
-••ft
•••ft
•••ft
••••ft
•••ft
‘•••ft
•••ft
•••ft
‘•••ft
‘•••ft
•••ft
‘•••ft
•••##!
••••ft
••••ft
‘•••ft
••••ft
I
•••ft
••••ft
•••ft
s*
-'T'TXI*Tr-222*;»I»*»** w »***»***»»*»»»f»fiftfffffffffffffffffffffffffffffff#fiffffftff
£•••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••
••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••••
MMuiuuuiiimmii ••••••••••••••••••••••••••••••••••••••••••••••••••••••••• ••##•
•••ft
••••ft
•••ft
••••ft
••••ft
••••ft
*••••<
••••ft
••••ft
••••ft
••••ft
••••ft
-•••ft
••••ft
••••ft
••••ft
•••ft
'*••••
•••ft#
••••ft
**••••
••••ft
••••ft
•••ft
••••#
•••ft
••••ft
-•••#
••••ft
•••ft
••••ft
••••ft
•••ft
••••ft
*••••#
••••ft
••••ft
••••ft
••••if
• • •_ w
;*»•»••
;••
ADVANCED
COMPUTER
ODUCTS
MOTOROLA
6800
COMPATIBLE
MODULES
MEK 6800 02 KIT $ 235.00
9600 6800 MPU 595.00
9601 16 Slot Mother Bd. 175.00
9602 1 6 Slot Card Cage 75.00
9603 8 Slot Mother Bd. 100.00
9610 Proto Board 36.00
9615 4K EPROM Module (1702A) 250.00
9620 16 Port Parallel I/O 375.00
9626 8K Static RAM Module 295.00
9626K 8K Static RAM Kit 225.00
9630 Extender Card 60.00
9650 8 Port Duplex Asyn Serial I/O 395.00
Connectors $6.50 6/84.00
All assembled & tested not Kits
PLUS MOTOROLA TV MONITORS-PRIME
Model M3560-155 L01 12" display 219.95
Model M 2000- 155 9" display 199.95
Add $10.00 for shipping
LOGOS 8K STATIC MEMORY
$ 125.99
Kit
Assembled 8< tested $179.95
250 ns. Kit 149.95
Assembled & tested 199.95
Features: Lowpower, Dip Switch
Selectable memory protect
down to 256 Bytes, addressing
“V v& c& on 1K Boundaries, No wait
^' e oe s' e ° states, fully buffered, battery
back-up.
SPECIAL OFFER:
Buy 4 Units Only $1 17.00 ea.
BYTEUSER 8K EPROM $ 64.95
Assembled 8t tested S 94.95
Bare PC Board w/Data 21.95
8K EPROM (8) 2708 $112.00
Features: Power on Jump, Reset
Jump, all socketed with top
quality PC Board material.
SPECIAL OFFER:
Buy 4 Units Only $59.95 ea.
Z-80 CPU KIT
Low Price
(regular price 269.95)
Assembled & tested
$ 129.95
Kit
$199.95
FEATURES:
S-100 IMSAI/Altair compatible,
completely compatible to TDL
hardware and software. Can be
used at 4MHz with Z-80A.
Add $7.00 for sockets.
Add $5.00 for Z80 manual.
NOW IN STOCK
FOR IMMEDIATE SHIPMENT.
DATA BOOKS
NSC Digital S 3.95
NSC Linear 4.95
NSC Linear A/N Vol. I 2.95
NSC Linear A/N Vol. II 2.95
NSC CMOS 2.95
NSC Audio 2.95
NSC Voltage Reg 2 95
NSC Memory 3.95
Intel Data Book 3.95
Intel MCS 85 Manual . 4 95
Intel MCS 80 Manual 7 95
Intel MCS 40 Manual 4.95
AMD 8080 Microproc. Handbook 7.95
AMI MOS Catalogue 3.95
Raytheon Linear . 2.50
Raytheon Quads/Duals . . 1.95
Gl MOS Catalogue 4 95
Osborne Intro to Micro Vol 0 . 7 50
Osborne Intro to Micro Vol I . 7.50
Osborne Intro to Micro Vol II . . 15.00
Osborne 8080 Programming 7.50
Osborne 6800 Programming 7.50
Osborne Z80 Programming ... 7.50
THE FIRST TO OFFER PRIME PRODUCTS TO THE HOBBYIST
AT FAIR PRICES NOW LOWERS PRICES EVEN FURTHER! A39
1. Proven Quality Factory tested products only, no re-tests
or fallouts. Guaranteed money back. We stand behind our products.
2. Same Day Shipment All prepaid orders with cashiers
check, money order or charge card will be shipped same day as received.
I MICROPROCESSORS
STATIC RAM HEADQUARTERS
AMI
6800 KIT
EVK 99
only
$ 133.00
Send for complete details
EVK 99 Kit ...
EVK 100 Kit . . .
EVK 200 Kit . . .
EVK 300 Assembled
Universal Kluge Board
16K Byte RAM Board
6 Slot Motherboard .
Extender Board . .
Video Board (avail, future)
Conrectors . . $6.50
Solid Frame Chassis .
Frame Chassis . . .
Micro Assembler ROM
Proto ROM ....
Tiny Basic Papertape
Tiny Basic EPROM .
$133.00
269.95
449.95
699.00
95.00
75.00
35.00
45.00
95.00
12/59.95
120.00
69.95
30.00
30.00
20.00
125.00
Hardware or Programming Manuals 15.00 ea
8080A CPU KIT $ 99.95
Kit
Assembled & tested $129.95
FEATURES:
S-100 bus compatible,
complete CPU with
eight level vector interrupt.
Includes sockets.
Add $5.00 for 8080 Manual
16K SUPERFAST RAM'S
pp D 416/4116 350ns. PRIME
Buy 8 pcs. — Only $29.95 each
Buy 16 pcs. — Only $27.95 each
Larger quantity or group buys
call for current price quote.
TARBELL FLOPPY INTERFACE
Complete Kit Only $1 79.95
Assembled & tested $269.95
FEATURES: SIOO/Altair/IMSAI
compatible, compatible
to most disc drives
including Persci, Innovex,
GSI, Shugart and others.
SOFTWARE: Uses CPM which is
available for $70.00
CPM documentation
add $20.00
Same Day Shipment
NAKED PC BOARD SALE S-100
Z 80 CPU $31.95
2708 EPROM 8K 21.95
8K Static RAM 21.95
32K Static RAM 59.95
Floppy I/O $39.95
Cassette I/O 29.95
Proto Bd. 27.95
Extender 15.95
UV “EPROM” ERASER
Model UVS-1 IE only $ 59.95
Holds (4) chips at a time.
Special holding tray with
UVA Absorber. Exclusive
safety interlock system.
Backed by 45 years UV experience.
WOW 1771 FLOPPY SPECIAL
LD While they last $32.95
O) regular price $55.95
CN Western digital P/N 1771A
^2 with App. Note & Data
Only 100 units available
•Z80
Z80A
F 8 13800)
2600
C01802
H080A
8085
8008 1
2901
2901 A
TMS 9900JL
CP 1600
6002
IM6100
• 6800P
6802P
1 30 1 20 1 15 •
?U4I«0«. I
4200A
~ ipD410 142001
10 90 10 50 9 90
29 95
14 90
21 90
29 90
89 95
39 95
19 95
29 95
19.95
32 90
8599
745189
745700
745201
A TO D CONVERTORS
SUPPORT DEVICES
• 3881 (Z80PIO) $12.95
14 95
350
995
3.75
4 85
12.95
2.25
395
795
995
11 95
21 95
• 3882 {Z80 CTC)
3851 (E8 Program)
3853 (F8 Memory I/O)
8212 8 bn I/O
8214 Priority Intenupl
8216 But dneer
8224 Clock Gen
8224 4 4MHe CLK Gen.
• 8T26 Buidriver . .
8226 Buidriver
8228 Syi Controller
8238 Svi Controller
8251/9551 Prog. Comm
8253 Interval T imcr
8255/9555 Prog Penph
8267 DMA Control
8259 Prog. Interrupt
6810 1 12B « 8 RAM
• 6820 PIA . . .
6834 1 EPROM 4K
6834 EPROM 4K
6850 ACIA
6852 Serial Adapter
6860 Modem 14 95
6862 RPS Modulator 17 95
• 687 IB 1.0 MH* Clock OSC .26.95
6880 MPU Busduver 2 95
68 Mini I Mmibug 2 49 95
68 Minlll Mimbug 3 74 95
6831/10224 Micro Aiiembler Rom 30 00
8700CJ
8701CN
8760CJ
1408L8
• 14816
S1395
21 95
1395
595
3.95
DISPLAYS/OPTO
7.95
14 95
16.95
995
Ol 704/707 CC/CA 300
FN0359CC 357
FND 500/507 CC/CA 500
FNO 503/510 CC/CA 500
FND 800/807 CC/CA 800
Bcwmar 9 digit bubble
FSC 8024 4 digit CC 800
HP7340 HEX Duplay
TIL 305 5. 7 Array
TIL 306 7 teg w/logic
TIL 308 7 teg w/logic
TIL 309 7 seg w /logic
TIL 311 HEX Duplay
MA 1003 12 auto clock
MA1002 4 digit clock module
MA1010 4 digit clock module
NSN 373/374 dual CC/CA 300
NSN 583/684 dual CC/CA 500
NSN 783/784 dual CC/CA 700
4N25 Opio Isolate/
MCT 2 Opto liolater
4N33 Darlington I SOL
Red Led'i 185 Dia
Green /Yellow
1 35
95
2 50
450
8.95
8.95
7.95
925
1795
995
9.95
2.20
260
3 00
3/1 99
89
1 75
LS163 1.39
LSI 64 139
LSI 74 1.39
LS17S 139
LSI 90 2 49
LS191 249
LS192 2.49
LS193 2.49
LS194 249
LS196 .52
LS2S1 1.39
LS253 1.49
LS257 1.29
LS2S8 1.29
LS279 1.10
LS283 1.10
LS367 99
LS36S 99
LS377 1 29
74200 5 96
74251 1 69
74279 1 10
74298 1 99
74365 79
74366 79
74367 79
74368 79
LS378 1.29
81LS96 1.10
81LS96 1 10
81LS97 1.10
81LS98 1.10
6831/1 1003 Proto Rom
1821SCD IK RAM
1822SCD 256 x 4 RAM
1824CD 32 . 8 RAM
1852CD 8 bit I/O
1B56CD I/O
1857CD I/O
30 00
15 95
1695
9 95
10 95
895
8 95
SOCKETS
22 Pm w
24 Pm w
40 Pin w
16 Pm S/T
18 Pm S/T
22 Pm S/T
24 Pm S/T
28 Pm S/T
40 Pm S/T
• 2708 $14.95
2708S 650ms 9 75
1702A 796
•1702 6 .
5204
• IMB610 . . .
6834
6834 1
82S123
82S126
82S129
8223
5203
Programming Available
CONNECTORS
3.50
16.95
2.95
16 95
296
295
695
6 Pm Single S/E
15/30 Dual S/E
18/36 Dual S/E
22/44 Dual S/E
43/86 Dual S/T
43/86 Dual W/W
50/100 IMSAI W/W
50/100 IMSAI S/T
50/100 Allan W/W
50/100 Actair S/T
MONTHLY SPECIALS
up 0371 Mag Tape Control $49 96
AYS 3550 4 3/4 Digit DVM 24 95
M K 5007/5009 Counter 6 95
AY5 3507 DVM Chip 12 96
ICM7208IPI Ctr /Disp/Dr.v 16 95
ICM7209 IPI Ctr/Drrver 19 95
I CM 7045 IPI Stop Watch 18 95
G115M 6 Channel MOS FETSW 150
4116/416 16K Dynamic RAM 34 95
Intel 3404 6 bit Latch 2 95
1488/1489 RS232 Dr.ver 125
8T97 BuHer 1.2S
8130/8131 2.45
8833/8835 1.99
74367/74368 6/5 00
75451/52/53 10/2 50
8T26 2/4.50
78L05 2/1.00
LM318H 100
4N25 Opto/Iso .69
SPECIAL 7/S1.00
7405 MPS6S30
7416 MPS3S68
7438 MPS6516
7440 MPS6S22
235
295
6.50
6.50
4.75
4 75
595
695
RAMS DYNAMIC
•4115
4116 16K
TMS4050
TMS4060 22 pm
W04060 22 pm
4096 16 nm
2104
2107
•2107B4
MM5261
MM 5262
MM5270
MM5280
1103
25 Pm 0 Submmatuie
DB25P 3.25
OB25S 3.75
SPECIAL 2/SI 00
7442 74123
7490 74145
7493 74153
7495 74LS195
IMSAI Card Guides
3.95
4 25
3 95
1 95
MCT2
2N3638A
2N3640
2N2369A
2N3692
S 89
5/1 00
5/1.00
5/1.00
5/1 00
MSPS4318 5/1 00
CRYSTALS
1.0MH, S 5.85 IO.OMH3 4 95
MPS5401
MPS6516
MPS6544
MPS3568
MPS6522
5/1 00
5/1 00
5/1 00
5/1 00
5/1 00
FLOPPY DISK
WO 1771 Floppy 32 95
WD 1771 H01 62 95
CHARACTER GENERATORS
R032513 Upper SV
R032513 Lower 5V
HO0165
MC6571A
MC6574
MC6575
S10 95
9 96
995
1080
14 50
2.0 585
2 097152 5.85
2.4576 585
3579545 3 95
40 4 95
4.194304 5.95
4.91520 595
4 95
5.95
4.95
50
5.7143
13.0 4 95
14 31818 4 95
18 0 495
18432 595
20 0 4 95
22 1184 5.95
27 000 5.95
36000 595
48000 595
100KC 12 95
•1PS3642 5 1 00
MPS6571 5/1 00
MPS3693 5/1 00
MPS6S30 5/1 00
MPS3646 5/1 00
752SJ 50
7524/ 60
74S258
7489
74145
74H55
715DC
74199
60
MPS404A 5/100
7N 5462 5/1 00
COMP KITS & SYSTEMS:
KEYBOARD ENCODERS
AYS2376 S14 95
AY53600 14 95
UARTS/USRTS
AY51013A 15V) . S 5 50
TR1602B 15V) 4 50
AY51014 IS 14VI 8 95
AYS101S 15V) 9 95
IM6402I6V) 12 95
IM6403I5V) 12 95
S2350 10 95
WD1671B Astros 19.95
BAUD RATE GENERATORS
NEW CTS DIPSWITCHES
CTS206 4 SI 75 CTS208 8 SI 95
CTS206 5 SI 75 CTS209 9 SI 95
CTS206 6 SI 75 CTS209 10 SI 95
CTS207 7 SI 75
MC14411
WD1941 Dual
34702
TV CHIPS
• MM5320 TV Synch . .
MM6369 Pn scale-
MM57 100 Game Chip
MM57104 Clock
LM 1889 Modulator
CW 300 Saw Function
AY 38500 1 TV Game
AY 38600 1 Color TV Game
AY38615 1 Color Converter
• AY 38700 1 Tank Chip .
RF Modulator
We also stock full of 7400,
74LS, 74L Linear and CMOS
Send for pricing or use our
competitors STD catalog
pricing.
NEW
1978 CATALOGUE
AVAILABLE FEB
SEND .25 POSTAGE
ME K6800D2 Kit
KIM1 6502
EVK99 6800 Kit
Techmco 9900 Kit
Intercept Jr 6100 Kit
lam 8080 Computet Book
NSC SC/MP Kit
NSC Keyboard Kit
Low Cost S 100 BUS with 8 i
& Power Supply
8K Ram Kit ILogosll
By tenser 8K Eprom Kit
ZPU Kit IZ80I
TDL ZPU Kn IZ80I
TarbeM cassette I/O
S100 8 slot Motherboard witl
connectors (expandable)
Si 00 Extender Board
S100 Proto Board
Vector 8800 Proto Board
IMSAI 8080 w/22 dot
Cromcmeo Z 2
TDL XITAN alpha 1
Com pu talker
Heuristics Speechlab
SOROC IQ1 20 Terminal
8K Ram Board (Logos)
8K Eprom Board
S100 32K Ram Kit
8ytesaver Kit
S235 00
245 00
133 00
299 00
281 00
499 00
99 00
95 00
its
159 95
125 95
64 95
295 00
269 00
115 00
69 95
15 95
27 95
19 95
699 00
595 00
769 00
385 00
299 00
995 00
21 95
21 95
875 00
145 00
74LS00TTL
$995
395
16 95
3 75
395
1 SO
995
24 95
895
. 29.95
895
ORGAN CHIPS
MM5S54 ]
MM5565 II each) „
MM 5556 1 S199&
CLOCK CHIPS
MM 53 14 S 4 9b
MM5316 4 95
MM5375 4 96
74LSOO
LS02
LS04
LS06
LS08
LS10
LS13
LS14
IS 20
LS27
LS30
LS32
LS40
LS42
LS51
LS73
LS74
LS75
LS76
LS83
LS85
LS86 55
LS90 110
LS92 1 10
LS93 1.10
LS95 1 69
LS107 56
LSI 09 56
LS112 56
LS1 13 .56
LS123 99
LSI 32 1.10
LS136 99
LSI 38 139
LS139 149
LS151 149
LS153 129
LS155 149
LS157 129
LS168 1.29
LS 161 139
LS162 139
NOTICE
We are looking for software
support packages to offer in
our new catalogue. Please
write or call if you want to
participate.
All Shipments FCM or UPS Ordi-ti
unde! SI 00 00 add 5 li.mdlenj and
IMSifuge Olden over S100 00 add 2 5
handling 8, postage Masteichatge Bank
anlenCard COO accepted sv 25 deposit
■HIS add 6 • tax Foieign
All p
P.O.BOX 17329 Irvine, California 92713 New Phone (714) 558-8813
TELEX/TWX: 910-595-1565
Retail Store Open Mon. - Sat.
Located at 1310 "B” E. Edinger,
Santa Ana, CA 92705
129
Beautiful Boards
OUR BEST SELLER: ECONORAM II™
S-100 Compatible 8K x 8 in a cost-effective package. Buffering
on all lines, 0 wait states with the 8080, low power consumption, con-
figured as two separate 4K blocks for addressing flexibility, handles
DMA, memory protect with vector interrupt provision if you try to
write into protected memory, fully socketed, gold-flashed edge
fingers, solder masked and legended board . . . this is the board that
doesn't cut any corners, but cuts the price instead.
(See the 1/77 issue of Kilobaud magazine for a product profile that
tells just about everything you’d ever want to know about Econoram II
... or send a self-addressed, stamped envelope to "Kilobaud Article”
do our address and we ll send you a reprint. But if you really want to
be convinced . . . talk to somebody who owns one!)
Kit form: $ 135.00 3 kits: $375.00 Assembled, tested: $ 1 55.00
ANNOUNCING . . . THE 16K ECONORAM IV T
We’ll be ready to ship these soon, so we thought you’d like a sneak preview. The
price? Under $400. The performance? All that you’ve come to expect from the
Econoram line, along with impressively low power consumption and a couple of
other tricks we have up our sleeve. If you’ve been waiting for a 16K board, you’ll be
happy you waited for us. ^
SOME WORDS ABOUT STATIC MEMORIES
When it comes to memory, we re pretty partial to static technology. Although more costly
than dynamic devices, static memories are free of critical refresh and timing needs — which
is one reason why DMA works so well with our memory boards. When we send an Econoram
out into the world, we not only want it to work right with whatever system you have (Allair.
IMSAI. Cromemco. Parasitic. Polymorphic, etc.): we want it to keep working for you. Static
memories are proven, time-tested, and reliable . . . that's why we like them so much.
ACTIVE TERMINATOR BOARD
The active termination circuitry in our motherboard kits minimizes the
ringing, crosstalk, overshoot, scrambled data, and noise problems that can
occur with unterminated lines. But even if you don't have a Godbout
motherboard, you can trick your computer into thinking you do by adding
this useful peripheral. Simply plug into any S-100 machine, and gain the
benefits of active circuitry. *CK-017, $29.50. Kit form only.
CPU POWER SUPPLY
Here is an economical supply for small computer systems or digital bench
work. Delivers 5V @ 4A with crowbar overvoltage protection (accidents can
happen . . . and you shouldn’t have to replace all your TTL if one does!).
Also gives + 12V @ VzfK and - 12V @ Viz A, along with an adjustable
negative bias supply ( - 5 to - 10V @ 10 mA). All in all, you can’t beat the
price or the performance. # CK-014, $50.00. Kit form only.
DB-25 RS 232 SUBMINID CONNECTORS
Male plug, *CK-1004, $3.95; female jack, *CK-1005, $3.95; plastic
hood for male connector, # CK-1006, $0.90. p— ■ — — 1
PLUG FROM BILL: There's more to life than computers . . . like music. Craig Anderton.
noted author and designer of our Musikit products, has produced a cassette tape of original
music that is distributed by our friends at PAIA Electronics (1020 W. Wilshire. Oklahoma
City, OK 731 16; $6.45 ppd). In addition to hearing our Musikits in action, you get to hear
some really good modern music. We like it . . . you probably will too.
TERMS: Please allow up to 5% for shipping; excess refunded. Californians add tax. COD
orders accepted with street address for (JPS. For VISA’ /Mastercharge* orders call our 24
hour order desk at (415) 562-0636. Prices good through cover month of magazine.
FREE FLYER: These are just a few of the items we carry for the computer enthusiast. We
also stock a broad line of semiconductors, passive components, and hobbyist items. We will
gladly send you a flyer describing our products upon receipt of your name and address.
SUPER MEMORY FOR A
SUPER MACHINE: H8 COMPATIBLE
ECONORAM VI™
Users of the S-100 buss have found out why our memories are
their best value . . . now H8 owners can find out too. This 12K x 8
kit offers the same basic features as our ECONORAM series . . . static
design, configuration as two blocks (one 8K and one 4K), switch
selected protect, sockets for all ICs, full buffering on address and data
lines . . . plus the required hardware and edge connector to mate
mechanically with the H8. As a bonus, all sockets and bypass
capacitors are pre-soldered to the circuit board so you can start right
in on the fun part of building this high-quality memory.
Kit form: $235.00
WE ALSO SPEAK DYNAMIC: ECONORAM III™
If you want a dynamic memory, might as well get one that works right.
Econoram III is inexpensive, completely assembled and tested, and ready to
plug into your S- 1 00 machine. Low power. 0 wait states with 8080 CPU. con-
figured as two 4K blocks, fully socketed.
$149.00, assembled and tested only.
EDGE CONNECTORS
There are edge connectors, and there are Edge Connectors. These are
the kind where the pins don’t fall out, thanks to the bifurcated contacts. (We
use the same connectors with our motherboards.)
*CK-1001: 100 pin edge connector with gold plated 3 level wrap posts.
Mates with Altair/IMSAI peripherals. $5 each or 5/$22.
# CK-1002: Same as above, but with soldertail pins on 0.25" centers.
(Mates with IMSAI motherboard). $5 each or 5/$22.
# CK-1003: Same as above, but with soldertail pins on 0.14" centers.
(Mates with Altair motherboard). $6 each or 5/$27.50.
10 SLOT MOTHERBOARD
Whether implemented as an add-on to existing systems that need more
room, or as the nucleus of a stand-alone system, this S-100 compatible
motherboard fits the needs of the budget-minded enthusiast. Our price in-
cludes all edge connectors, along with active termination circuitry that pro-
motes accurate and reliable data transfer. Lots of bypass caps and extra
heavy power line traces contribute to efficient operation. Heavy duty epoxy
glass board, with a solder mask for easy soldering.
*CK-015, $90.00. Kit form only.
18 SLOT MOTHERBOARD
All the same features and advantages of the 1 0 slot version, including our
active termination circuitry. Complete with 18 edge connectors.
*CK-016, $124.00. Kit form only.
DEALER NOTE
We’d like to thank the ever-growing number of dealers who are spreading the
Econoram word to their customers . . . you will be happy to know that we have
doubled the capacity of our Compukit 1 " division in order to continue handling the
massive response. We re glad you like what we re doing . . . and we’re going to keep
on doing it!
BILL GODBOUT ELECTRONICS
BOX 2355. OAKLAND AIRPORT. CA 94614
G4
130
Hifioim itm
600 MHZ. FREQUENCY COUNTER
±0.1 PPM TCXO
VHF RF Pick-Up Antenna-Rubber Duck w/BNC #Duck-4H $12.50
Right Angle BNC adapter #RA-BNC $ 2.95
FC-50 — Opto-8000 Conversion Kits:
Owners of FC-50 counters with #PSL-650 Prescaler can use
this kit to convert their units to the Opto-8000 style case, includ-
ing most of the features.
FC-50 — Opto-8000 Kit $59.95
* FC-50 — Opto-8000F Factory Update $99.95
FC-50 — Opto-8000. 1 (w/TCXO) Kit $109.95
* FC-50 — Opto-8000. IF Factory Update $149.95
* Units returned for factory update must be completely as-
sembled and operational
TERMS: Orders to U.S. and Canada, add 5% to maximum of $10.00 per order
for shipping, handling and insurance. To all other countries, add 10% of total or-
der. Florida residents add 4% state tax. C.O.D. fee: $1.00. Personal checks must
clear before merchandise is shipped.
This new instrument has taken a giant step in
front of the multitude of counters now available.
The Opto-8000. 1 boasts a combination of fea-
tures and specifications not found in units cost-
ing several times its price. Accuracy of ±0.1
PPM or better — Guaranteed — with a
factory-adjusted, sealed TCXO (Temperature
Compensated Xtal Oscillator). Even kits re-
quire no adjustment for guaranteed accu-
racy! Built-in, selectable-step attenuator, rug-
ged and attractive, black anodized aluminum
case (.090" thick aluminum) with tilt bail. 50
Ohm and 1 Megohm inputs, both with
circuits for super sensitivity and both
diode/overload protected. Front panel in-
cludes “Lead Zero Blanking Control” and a
gate period indicator LED. AC and DC
power cords with plugs included.
ACCESSORIES:
Battery-Pack Option — Internal Ni-Cad Batteries and charging unit
$19.95
Probes: P-100 — DC Probe, may also be used with scope $13.95
P-101 — LO-Pass Probe, very useful at audio frequencies
$16.95
P-102 — High Impedence Probe, ideal general purpose
usage $16.95
SPECIFICATIONS:
Time Base— TCXO ±0.1 PPM GUARANTEED!
Frequency Range — 10 Hz to 600 MHz
Resolution — 1 Hz to 60 MHz; 10 Hz to 600 MHz
Decimal Point — Automatic
All IC’s socketed (kits and factory-wired)
Display — 8 digit LED
Gate Times — 1 second and 1/10 second
Selectable Input Attenuation — XI, X10, X100
Input Connectors Type — BNC
Approximate Size — 3"h x 7V2"w x 6V2"d
Approximate Weight — 2V2 pounds
Cabinet — black anodized aluminum (.090" thickness)
Input Power— 9-15 VDC, 115 VAC 50/60 Hz
or internal batteries
OPTO-8000. 1 Factory Wired
OPTO-8000. IK Kit
$299.95
$249.95
OPTOELECTRONICS. INC.
5821 NE 14 Avenue
Ft. Lauderdale, FL 33334
Phones: (305) 771-2050 771-2051
Phone orders accepted 6 days, until 7 p.m. Q3
7400 TTL
SN7400N
SN7401N
SN7402N
SN7403N
SN7404N
SN7405N
SN7406N
SN7407N
SN7408N
SN7409N
SN7410N
SN7411N
SN7412N
SN7413N
SN7414N
SN7416N
SN7417N
SN7420N
SN7421N
SN7422N
SN7423N
SN7425N
SN7426N
SN7427N
SN7429N
5N7430N
SN7432N
SN7437N
SN7438N
SN7439N
SN7440N
SN7441N
SN7442N
SN7443N
SN7444N
SN7445N
SN7446N
SN7447N
SN7448N
SN7450N
SN7451N
SN7453N
SN7454N
SN74S9A
SN7460N
SN7470N
SN7472N
SN7473N
SN7474N
SN7475N
SN7476N
SN7479N
SN7480N
SN7482N
SN7483N
SN7485N
SN7486N
SN7489N
SN7490N
SN7491N
SN7492N
SN7493N
SN7494N
SN7495N
SN7496N
SN7497N
SN7410ON
SN74107N
SN74109N
SN74116N
SN74121N
SN74122N
SN74123N
SN74125N
SN74126N
SN74132N
SN74136N
SN74141N
SN74142N
SN74143N
SN74144N
SN74145N
SN74147N
SN74148N
SN74150N
SN74151N
SN74153N
SN741MN
SN74155N
SN74156N
SN74157N
2.95
2 95
2.95
SN74160N
SN74161N
SN74162N
SN74163N
SN74164N
SN74165N
SN74166N
SN74167N
SN74170N
SN74172N
SN74173N
SN74174N
SN74175N
SN74176N
SN74177N
SN74179N
SN74180N
SN74181N
SN74182N
SN74184N
SN74185N
SN74186N
SN74188N
SN74190N
SN74191N
SN74192N
SN74193N
SN74194N
SN74195N
SN74196N
SN/4197N
SN74198N
SN74199N
SN74200N
SN74251N
SN74279N
SN74283N
SN74284N
SN74285N
SN74365M
SN74366N
SN74367N
SN74368N
SN/4390N
SN74393N
“TOTT
CD4001
CD4002
CD4006
CD4007
CD4009
C04010
C04011
CD4012
C04013
CD4014
CD4015
CD4016
CD4017
CD4018
CD4019
CD4020
CD4021
C04022
C04023
CD4024
CD4025
CD4026
C04027
CD4028
CD4029
CD4030
CD4035
CD4040
C04041
CO4042
004044
CD4046
CD4047
20% Discount for 100 Combined 7400's
C M OS
CD4049 .45
1 19
2.95
9.95
1.19
.49
1.19
C04050
CD405I
C04053
C04056
C04059
C04060
C04066
CD4068
C 04069
CC4070
CD4071
C04072
CD4076
CC4081
C 04082
CD4093
CD4098
MCI 4409
MC14410 14.95
78VG
LM3O0H
LM301H
LM301CN
LM302H
LM304H
LM305H
LM307CN/H
LM308H
IM308CN
LM309H
LM309K
LM310CN
LM311H
LM311N
LM317K
LM318CN
LM319N
LM320K-5
LM320K-5.2
LM320K-12
LM320K-15
LM320K-18
LM320K-24
LM320T-5
LM320T-52
LM320T-8
LM320T-12
LM320T-15
LM320T-18
LM320T-24
LM323K-5
LM324N
LM339N
LM340K-5
LM340K-6
LM340K-8
LM340K-12
LM340K-15
LM340K-18
LM340K-24
LM340T-5
LM340T-6
1.79
2 50
-L35-
MC 14411
MC14419
MCI 4433
MCI 4506
MCI 4507
CD4508
C04510
C04511
CD4515
C04518
CO4520
MC14562
C04566
14.95
4 95
14.95
74C02
74C04
74C08
74C10
74C14
74C20
74C30
74C42
74C43
74C73
74C74
74C89
74C90
74C93
74C95
74C107
74C151
74C154
74C157
74C160
74C161
74C163
74C164
74C173
74C192
74C193
74C195
74C922
74C923
74C925
74C926
3.00
2.00
2.00
1.25
290
3.00
2.15
3.25
3.25
3.00
3.25
2 SO
3.49
2.75
2 75
9.95
1.50
1.10
1 25
1.15
1.25
1.25
1.25
LINEAR
LM340T-8 125
LM340T 12 125
LM340T-15 1-25
LM340T-18 1.25
LM340T -24 125
LM350N 100
LM351CN
LM370N
LM373N
LM377N
LM380N
LM380CN
LM381N
LM382N
NE5D1N
NE510A
NE529A
NE531H
NE536T
NE5401
NE550N
NE555V
NE556
NE560B
NE561B
NE562B
NE565H
NE555N
NE566CN
NE557H
NE557V
NE570
LM703CN/H
LM709H .29
LM709N .29
LM710N .79
LM711N .39
LM723H .55
LM723N 55
1.15
3.25
4.00
1.25
99
1.79
1.79
8.00
6.00
4 95
1.19
LM733N
I.M739N
LM741CH
LM741CN
LM741-14N
LM747H
LM747N
LM748H
LM748N
LM1303N
LM1304N
LM1305N
LM1307N .85
LM1310N 2.95
LM1351N 165
LM1414N 175
LM1458CN/H .59
MCI 488 1.95
MCI 489 1.95
LM1496N .95
LM1S56V 1.75
MC1741SCP 3.00
LM2901N 2.95
LM3053 1.50
LM3065N 69
LM3900N(3401) .49
LM3905N 89
LM3909 1.25
MC5558V 1.00
LM7525N .90
LM7534N
10.50
LM754S0
75451 CN
75452CN
75453CN
75454CN
75491CN
75492CN
75494CN
RC4151
RC4194
RC4195
4.95
.49
74LSOO
74LS02
74LS03
74LS04
74LS05
74LS08
74LS10
74LS13
74LS14
74LS20
74LS26
74LS27
741S28
74LS30
74LS32
74LS40
74LS42
74LS47
74LS51
74LS55
k 74 IS 73
74LS00 TTL
74LS74
74LS75
74LS76
74LS83
74LS85
74LS86
74LS90
74LS92
74lS93
74LS95
74lS96
74LS107
74LS109
74 LSI 12
74LS123
74LS132
74LS136
741S138
74LS139
74LS151
74LS155
74LS157
74LS160
74LS161
74LS162
74LS163
74LS164
74LS175
74LS181
74LS190
74LS191
74LS192
74LS193
74LS194
74LS195
74LS253
74LS257
74LS260
74LS279
74LS367
74LS368
74LS670
BUGBOOK-
Continuing Education Senes
se( these two books outline over SO eiperiments designed
reader al he »na need to know about TTL logic claps ic use then
| unction with microprocessor systems You'll team about the basic concepts of
digital electronics including gates, I s flops, alches buses, decoders, muifi-
plesers. demu tlplexers. LEO displays. RAM's. ROM's, and much, much mote
$5.00
interface between data terminals, etc . and your microcomputer It a<so covers
currant loops, and the RS 232C inteiace standard Particularly recommended
for any RTTY enthusiast
$6.95 BUGBOOK III $i05
by Peter R Reny. Oenfd 6. Linen. W84KYJ. Jonathan A. Titus
THE 555 TIMER APPUCATIONS
SOURCEBOOK WITH EXPERIMENTS
by Howard M. Berlin W3HB
This book shows you what the 555 tmer is and how to use it Included are over
100 various desgn techniques, equations and graphs to create ready-to-go" 8080 chip pm by pin and introduces you to the Mark SO microcomputer,
Umars, generators, power supples, measure men l and control emeu its, party unique easily interfaced system It is recommended that you have the back-
games. circuits tor the none and automobile, photography, music and ground on the BUGBOCKS I & ll before proceeding wdh BUGBOOK
mamma ! sTnfi BUGBOOK V and VI $19.00 par tel
INSTRUCTOR S MANUAL S3. 00 bf omuld G. Larsen. Peter fl Rony, Jonathan A Tlltii
Necessary lor instruction of Bugtoo* I ark) II. Answers questions regarding . . .
experiments, suggestions far tjrtter read: vg. philosophy of authors apprwch to !S? nmenlJ ,n ^ microcomputer programming and
digital electro lies A must tor self-teaching Individuals. «080A microcomputer interfacing An integrated approach to se/fmstaefed
bas’C dogita electronics, breadboards ard 8080A interlacing .'orograivn ig
‘ s tne dogiu
:er prograr
w.lh each
OP AMP MANUAL by Howard M. Berlin W3H0 $9.00 Bug^vi mtec^ tne ^w concip." oi Bugbcok
expenment guide to appflcaton of operational ampMlers Over 25 expen- microcomputer programming and interlacing Detail & laborato-y
fits on all phases of Op Amps. * ** ‘” K
$O0 CMOS-M — DESIGNERS PRIMER SC
8080 interpretive debugger A program fer entenng. debugging and storing AND HANDBOOK
COMPLETE MANUAL FOR OIGITAL CLOCKS by John Weiss and John Brooks
Familiarizes technician or hobbyist with basic theories behind digital clocks, includes trouble shooting guides, basic
characteristics o' docks, soldering techniques, clock component data sheets and construction lips. $3.95
XC209
XC209
XC209
XC209
XC22
XC22
XC22
XC22
SSL -22
125" dia.
Red 5/$1
Green 4/$i
Orange 4/Si
Yellow 4/SI
.200" dia
Red 5/SI
Green 4/SI
Yellow 4/51
Orange
DISCRETE LEDS
RT
4/SI
XC526
XC526
XC526
XC526
XC526
Green
Yellow
Clear
XC556
XC556
XC556
XC556
XC556
XC556
Red
Red
Green
Yellow
Orange
Clear
XC111
XC111
XC111
XC111
5/51
4/SI
7/51
.190" dia.
Red 10/S1
Green 4/SI
Yellow 4/Si
Orange 4/$1
.085" dia.
MV50 - Red • 6/SI
inFM-Aed'Ted"
V x V«" x 1/16
Rat j ffi. qq .
DISPLAY LEDS
MAN 52
MAN 71
MAN 72
MAN 74
MAN 81
MAN 82
MAN 84
MAN 3620
MAN 3630
MAN 3640
MAN 4610
MAN 4640
MAN 4710
MAN 4730
MAN 4740
MAN 4810
MAN 6610
MAN 6630
POLARITY
Common Anode-red
5 x 7 Dot Matrix -red
Common Cathode-red
Common Cathode-red
Common Anode-green
Common Anode-red
Common Anode-red
Common Cathode-red
Common Anode-yellow
Common Anode-yellow
Common Cathode-yellow
Common Anode -orange
Common Anode-orange *1
Common Cathode -orange
Common Anode -orange
Common Cathode-orange
Common Anode-red si
Common Anode-red
Common Cathode-red
Common Anode-yellow
Common Anode-orange-D
Common Anode -orange
Common Cathode-orange-
Common Cathode -orange
Common Miode-oranye
PRICE
2.95
4.95
.560
•O.D. 560
tl 560
TYPE
MAN 6680
MAN 6710
MAN 6730
MAN 6740
MAN 6750
MAN 6760
MAN 6780
OL701
0L702
DL704
DL707
OL741
0L746
DL747
OL749
DL750
DL33B
FND70
FN0359
FND503
FND507
5082-7300
5082-7302
5082-7304
5082-7340
POLARITY
Common Cathode -orange
Common Anode-red-D.D.
Common Anode-red ? 1
Common Cathode-red -D O
Common Cathode-red ±1
Common Anode-red
Common Cathode-red
Common Anode-red -1
Common Ca!hode-red
Common Cathode -red
Common Anode-ned
Common Anode-red
Common Anode-red ±1
Common Anode-red
Common Cathode-red ±1
Common Cathode-red
Common Cathode-red
Common Cathode
Common Anode
Common Cathode (FN3500)
Common Anode (FN0510)
4x7 Sgl. Oigit-RHOP
4x7 Sgl. Diglt-LHDP
Overrange character (±1)
4 x 7 Sgl. Digit-Hexadecimal
.560 . 99
300 .99
300 1.25
.600 19.95
.600 15.00
600 22.50
RCA LINEAR
CA3013
CA3023
CA3035
CA3039
CA3046
CA3053
CA3059
CA3060
CA3080
CA3081
CA3Q82
2 15 C A3 083
2.56 CA3086
2.48 CA3089
135 CA3091
1 30 CA3102
,50 CA3123
3 25 CA3130
3 25 CA3140
85 CA3160
2 oo CA3401
2QQ CA360Q
3 75
3.50
2.95
2.15
1.39
XR-2206KB Kit $19.95
WAVEFORM
GENERATORS
XR 205 $8.40
XR-2206CP 5.50
XR-2207CP 3.85
EXAR
STEREO DECOOERS
XR-1310CP $3.20
XR-1800P 320
XR-2567 299
XR-2211CP
XR-4136
XR 1468
XR-1488’
XR-1489
XR-2208
XR-2206KA Kit $14.95
TIMERS
XR-555CP $.39
XR-320P 1.55
XR-556CP
XR-2556CP 3.20
XR-2240CP 4.80
PHASE LOCKED LOOPS
XR-210 520
XR-215 6 60
XR-567CP
XR-5B7CT 1.25
1-24
8 pm LP S 17
14 pm LP 20
16 pm LP 22
18 pm LP 29
20 pin LP .34
14 pm ST $ 27
16 pm ST 30
18 pm ST 35
8 pm SG S 30
14 pm SG 35
16 pm SG 38
18 pm SG 52
8 pin WW $.40
10 pm WW 45
14 pm WW 39
16 pm WW 43
» pin WW 75
IC SOLDERTAIL — LOW PROFILE (TIN) SOCKETS
50-100 > 1-24
flUMk 22 pm LP $ 37
MV 24 pm 38
20 •■■PH r I " 28 pm LP 45
27 36 pm LP 60
30 SOLDERTAIL STANDARD (TIN) « O'" LP 63
28 pm ST S 99
36 pm ST 1 39
40 pin ST 1 59
SOLDERTAIL S TANDARD (GOLD)
24 pm SG S 7C
28 pm SG 1 10
« — •— — 36 pm SG 1 75
43 40 pm SG 1 75
WIRE WRAP SOCKETS (GOLD) LEVEL #3
.35 ■■■ 22 pin WW %
24 pm WW 1 05
37 28 pin WW 1 40
1 10
1 30
1 40
50 PCS.
ASST. 1
ASST. 2
ASST. 3
ASST. 4
ASST. 5
ASST. 6
ASST. 7
ASST. 8R
RESISTOR ASSORTMENTS $1 .75
10 OHM 12 OHM 15 OHM 18 OHM 22 OHM
5 ea. 27 OHM 33 OHM 39 OHM 47 OHM 56 OHM 1/4
68 OHM 82 OHM 100 OHM 120 OHM 150 OHM
Sea. 180 OHM 220 OHM 270 OHM 330 OHM 390 OHM
470 OHM 560 OHM 680 OHM 820 OHM IK
5 ea. 1 2K 1.5K 1 8K 2 2K 2.7K
4 7K 5.6K 6 8K
3.3K
22K
150K
S ea. 390K
3.9K
10K
27K
60K
180K
470K
1.2M
3 3M
12K
33K
220K
560K
15K
39K
100K
270K
680K
18K
47K
120K
330K
820K
2 2M
Includes Resistor Assortments 1 -7 (350 PCS.)
PER ASST.
4 WATT 5% 50 PCS.
4 WATT 5% 50 PCS
4 WATT 5% 50 PCS. ^
4 WATT 5% = 50 PCS.
4 WATT 5% 50 PCS.
4 WATT 5% = 50 PCS.
4 WATT 5% = 50 PCS.
$9.95 ea.
$5.00 Minimum Order — U S. Fundi Only
California Residents - Add 6% Sales Tax
Spec Sheets - 2Se — Send 35c Stamp for 1978 Catalog
Dealer Information Available
ELECTRONICS
1021 -A HOWARD AVE., SAN CARLOS. CA. 94070
PHONE ORDERS WELCOME — (415) 592-8097
Advertised Prices Good Thru May
WIRE-WRAP KIT — WK-2-W
WRAP • STRIP • UNWRAP
• Tool for 30 AWG Wire
• Roll of 50 Ft. White or Blue 30 AWG Wire^
• 50 pcs. each 1", 2". 3" & 4" lengths —
pre-stripped wire.
$12.95
WIRE WRAP TOOL WSU-30
WRAP • STRIP • UNWRAP $6.95
WIRE WRAP WIRE — 30 AWG
25ft. min. $1.25 50ft. $1.95 100ft. $2.95 1000ft ..$15.00
SPECIFY COLOR — White - Yellow - Red - Green - Blue - Black
WIRE DISPENSER - WD-30
• 50 ft. roll 30 AWG KYNAR wire wrap wire $3.95 ea.
• Cuts wire to desired length
• Strips 1’’ of insulation Specify — Blue-Yellow-White-Ret
REPLACEMENT DISPENSER SPOOLS FOR WO 30
Specify blue, yellow, white or red SI .98/spool
Prime
Inventory
p/n
SW7401
SW7407
SW7416
SW7417
SW7420
SW7423
SW7425
SW7427
SW7430
SW7432
SW7437
SW7438
SW7440
SW7443
SW7444
SW7445
SW7446
SW7450
SW7453
SW7454
SW7460
SW7472
SW7475
Clearance STEWART WARNER S ^' r al
iln.l
19.00
19.00
19.00
11.00
19.0Q
1400
12.00
12.00
12.00
12.00
V748? S4.90/1ot 44.00/lot 430.00
180.00 SW7483 3.50
180.00 SW7486 2.20
180.00 SW7491 3.50
100.00 SW7494 3.50
18000 SW7495 3.50
130.00 SW7496 3.50
110.00 SW74100 6 90
110.00 SW74104 2.20
110.00 SW74105 2.20
110.00 SW74107 2.20
110.00 SW74121 2.20
80.00 SW74123 2.50
300.00 SW74145 5.50
300.00 SW74150 6.90
300 00 SW74151 3.50
350.00 SW74153 4.00 36 00
80.00 SW74156 4.00
80.00 SW74180 4.50
80.00 SW74181 9.90
80.00 SW74182 4.50 41.00
180 00 SW9601 2.50 22.00
300.00 SW9602 4.90 44.00
31.00
19.00
31.00
31.00
3100
31.00
66.00
19.00
19.00
19.00
19.00
22.00
51.00
66 00
31.00
41.00
30000
180.00
300 00
300.00
300 00
300.00
650.00
180.00
180.00
180.00
180.00
200.00
900.00
650.00
300.00
350.00
350.00
400.00
SW7480 2.90 26 00 250.00
Pre-tubed • No mixing or combining prices
TV GAME CHIP SET — $7.95
Includes AY -3-8500-1 Chip and 2.010 mhz crystal
(2.010 crystal — $.99 ea/AY-3-8500-t Chip — $7,50 ea.)
TYPE
1N746
1N751A
1N752
1N753
1N754
1N5232
1N5234
1N5235
1N5236
1N456
1N458
1N485A
1N4001
IN4002
1N40O3
1N4004
ZENERS -
VOLTS W
3.3 400m
5.1 400m
5.6 400m
6.2 400m
6 8 400m
6.2 40Om
15 40Om
5.6 500m
6.2 5flOm
6.8 50Om
7.5 500m
- DIODES —
PRICE TYPE
RECTIFIERS
VOLTS W
600 PIV 1 AMP
800 PIV 1 AMP
1000 PIV 1 AMP
200m
10m
10m
7o|
iom
28 1N4735
28 1N4736
28 1N4738
6/1.00 1N4742
50 PIV 1 A
100 PIV 1 AMP
200 PIV 1 AMP
400 PIV 1 AMP
12
15
l 50 PIV 35 AMP
100 PIV 35 AMP
i 150 PIV 35 AMP
i 200 PIV 35 AMP
i 400 PIV 35 AMP
PRICE
10/1.00
10/1.00
10/1.00
6/1.00
15/1.00
12/1.00
20/1.00
1.70
1.50
1.80
3.00
SCR AND FW BRIDGE RECTIFIERS
C360 15A@400V SCR $195
C38M 35A @ 600V SCR 195
2N2328 1 ,6A @ 200V SCR .50
MDA 980-1 12A@50V FW BRIDGE REC. 1.95
MPA 980-3 1 2A @ 200V FW BRIOGE REC. 1.95
MPS A06 3/SI. 00
2N918 4/31.00
2N2219A 3/31.00
2N2221 4/31.00
2N2222* 5/31,00
2N2369 5/31.00
2N2369A 4/3100
2N24B4 4/31.00
2N2906A 4/31.00
2N2907A 5/31.00
5/31.00
2/3100
3 M
TRANSISTORS
?
2N3055
2N3392
3125
3100
S/3100
5/31 OO
PN356T 3/3100
PN3568 4/31.00
1*113569 4/31.00
2N3702
2143704
2N3705
21/3706
24,3707
244371 1
2N3724
2N4014
2N4123
PN4249
5/31 00
5/31 00
5/31.00
5/31.00
1
PN4250 4/31.00
2414400 4/31.00
254401 4/31.00
24/4402 4/31.00
2K4403 4/S1.00
254409 5/31 30
4/31.00
4.31 .00
431.00
531 00
531.00
531.00
531.00
S/51.00
32 00
331 OO
53100
2N5129
2*45138
255139
2N5209
2N5210
2N5432
C1Q6S1SCR 2/3100
40439 *1.75
40413 51. 75
CAPACITOR
50 VOLT CERAMIC CORNER
OISC CAPACITORS ^ 1 1
1-9
10-49
50-100
1-9
10-49
50-100
10 pi
m
.04
.03
.001#4F
.05
.04
.035
22 pf
05
.04
.03
.0047*iF
.05
.04
.035
47 pf
09
.04
.03
,01/iF
09
.04
.035
100 pf
05
.04
.03
. 022/u F
.06
.05
.04
220 pf
05
M
.03
047 m F
.06
.05
.04
470 pf
.05
.04
035
IuF
.12
.09
.075
100 VOLT MYLAR FILM CAPACITORS
.OOlmf
.12
.10
.07
022m!
.13
.11
00
.0022
0047mf
.12
10
07
,047mf
.21
.17
.13
.12
10
.07
.imt
.27
.23
.17
Olmf
12
.10
07
,22mt
.33
.27
.22
+20% OIPPED TANTALUMS (SOLID) CAPACITORS
1/35V
28
.23
.17
1.5/35V
.30
.26
.21
15/35V
.28
.23
.17
2.2/25V
.31
.27
22
22/35 V
23
.23
.17
3.3/25V
.31
27
.22
.33735V
28
.23
.17
4.7/25V
.32
.20
.23
•47/35V
28
.23
.17
6.8725V
.36
31
.25
68735V
28
.23
.17
10/25V
.40
35
.29
1.0/35V
.28
.23
.17
15/25V
.63
.50
.40
MINIATURE ALUMINUM ELECTROLYTIC CAPACITORS
Axial Lead
.47/50V .15 .13
1 0750V .16 .14
3.3/SOV .14 .13
4.7/25V .16 14
10/25V
.16
10750V
22725V
22/5 OV
47/25V
47/50V .25
100/25V .24
100/50V .35
220/25V .32
220/50V .45
470/25V ”33
1000716V .55
2200716V .70
Radial Lead
47/25V .15 .13
47750V 16 .14
1.0/16V .15 .13
1.0/25V .16 .14
1.0/50V .16 .14
4.7/16V .15 .13
4.7/25V .15 .13
4.7/50V .16 .14
10/16V .14 .12
10725V .15 .13
10/50V .16 .14
47/50V .24 .21
100/16V .19 .15
100/25V .24 .20
100/50V .35 .30
220716V .23 17
470/25V .31 .28
J
132
/Is
JUMPERS
• u i a . izubih un pdiiems ui tuu
centers and shielded receptacles.
Probe access holes in back. Choice
of 6" or 18"
length.
Part No. No. of Contacts
Length
Price
924003 -18R
26
18"
$ 5.38 ea.
924003 -06R
26
6"
4.78 ea.
924005 -18R
40 •
18"
8.27 ea.
924005 -06R
40
6"
7.33 ea.
924006-18R
50
18"
10.31 ea.
^24006-pflfl
50
6"
9.15 ea.
II SOCKET
Mates with two rows of .025" sq. or
IIMPFR Solder t0 PC boards for instant
HI wlilfll tn piug-in access via socket -connector
HEADERS
jumpers. .025" sq. posts. Choice
of straight or right angle.
Part No. No. of Posts
Angie
Price
923863 -R
26
straight
$1.28 ea.
923873 -R
26
right angle
1.52 ea.
923865-R
40
straight
1.94 ea.
923875 -R
40
right angle
2.30 ea.
923866 -R
50
straight
2.36 ea.
923876-R
50
right angle
2.82 ea.
IS INTRA-CONNECTOR
Provides both straight and right angle functions. Mates
with standard .10" x .10" dual row connectors (i.e. 3m, Ainsley,
etc.) Permits quick testing of inaccessible lines.
Part No.: 922576-26 No. of contacts: 26 Price $6.90 ea.
I! INTRA-SWITCH
Permits instant line-by-line switching for diagnostic or QA
testing. Switches actuated with pencil or probe tip. Mates with
standard .10" x .10" dual-row connectors. Low profile design.
Switch buttons recessed to eliminate accidental switching.
Pari No,: IS-26 No. of contacts. 26 Price $13 80 ea
' W CRYSTALS 3T-
— THESE FREQUENCIES ONLY
THESE FREQUENCIES ONLY 5
-u“
PirtV
Frequency
Caie/Style
— Prfcf
CY1A
1 000 MHz
HC33U
$595
CY2A
2 00C MHz
HC33U
S5.95
ICY2 01
2.010 MHz
HC33/U
OH
CY3A
4 000 MHz
HC18/U
$4 95
CY7A
5 000 MHz
HC18U
S4 95
CY12A
10 000 MHz
HC18U
$4 95
CY14A
14 31818 MHz
HC18U
S4 95
CY19A
18 000 MHz
HC18U
S4 95
CY22A
20 OOO MHz
HC18U
S4 95
CY30B
32 000 MHz
HC18U
S4 95
CONNECTORS
PRINTED CIRCUIT EDGE-CARD
.156 Spacing-Tin-Double Read-Out
Bifurcated Contacts — Fits .054 to 070 P.C. Cards
15/30
PINS (Solder Eyelet)
$1.95
18/36 1
PINS (Solder Eyelet)
S2.49
22/44 >
PINS (Solder Eyelet)
$2.95
50/1 00A ( ioo SMcmq) PINS (Wire Wrap)
25 PIN-D SUBMINATURE (RS232)
$6.95
DB25P
PLUG
S3. 25
DB25S
SOCKET
S4.95
DB51 226-1
COVER FOR 25S/25P
$1.75
siis
ers *
LOTS OF POTS £
J Untested W square Spectrol Trimpots'
"Single-turn Printed Circuit Potentiometers
GB134 3 ea. of: ] 00-200-250-500 ohm = 24 pcs $2.95
GB135 3 ea. of: = 24 pcs $2.95
gbi 36 -»*“■ * 2 - 95
(Values subject to substitution within each group.)
EXTRA SAVINGS’ Buy all 3 (GB134, 135 8 136) for only$7.49
%" mounting holes
•?. TOGGLE
RS*! (sub-mmature)
SWI TCHES
JMT121 SPOT on-ott-on
JMT123 SPDT on-none-on
JMT221 OPDT on-off -on
JMT223 DPDT on-none-on
1 58
MPC121 SPDT on-off-on $2 05 $153
TOGGLE M PCI 23 SPOT on-none-on 175 131
(Printed Circuit) MPC221 DPOT on-off-on 2 65 1 97
MPC223 DPDT on-none-on 2 25 1 68
II
|t PUSH BUTTON
PB123
PB126
SPDT maintained 195
SPDT momentary 1 95
II
PUSH BUTTON MS102 DPST momentary open 35 30
Mmature MS103 SPST momentary closed .35 30
206-4 8 pm dip 4 switch 1.75 165
206-7 14 pin dip 7 switch 195 1 85
206-8 16 pin dip 8 switch 2 25 2 15
1/16 VECTOR BOARD
1. 1 Hole Spacing P-Pa
64P44 062XXXP
169P44 062XXXP
64P44 062WE
84P44 062WE
169P44 062WE
169P84 062WE
169P44 062WEC1
4 50
450
4.50
4.50
I 50
4 50
W 1-9 10 up
6 50 1 72 1 54
4 53
8 26
6.12
INSTRUMENT/
CLOCK CASE
Injection molded unit.
Complete with red bezel
4VV x 4" x 1-9/16"
$3.49
MICROPROCESSOR COMPONENTS'
8080A
CPU
S10.95
Z80
CPU
$24.95
8212
8 Bit Input/Output
4.95
CD PI 802
CPU
19.95
8214
Priority Interrupt Control
795
MC6800
8 Bit MPU
19.95
8216
Bi-Directional Bus Driver
4.95
MC6810AP1
128 x 8 Static RAM
5.95
8224
Clock Generator/Driver
5.95
MC6820
Periph. Interface Adapter
7.95
8228
System Controller Bus Driver 5.95
MC6830L8
1024 x 8 Bit ROM
14.95
CPU'S
RAMS
0O8OA
Super 8008
S 10.95
1101
256 i 1 Static
S 1 4
2650
8 Bit MPU
26 SO
2101
256 x 4 Static
5.1
P8065
CPU
29 95
2102
1024 x 1 Static
1.1
2504
2518
2519
2522
2524
2525
2527
2528
2529
2532
2533
3341
74LS670
SR'S
1024 Dynamic
Hex 32 BIT
Hex 40 BIT
Dual 132 Bit SSR
512 Dynamic
1024 Dynamic
Dual 256 BIT
Dual 250 Bit Static
Dual 512 BIT
Quad 60 BIT
1024 State
Fifo
$395
4 95
4 00
295
99
2.95
2.95
400
400
2.95
2.95
6 95
1.95
210715280 4096
2111 256 x
TMS4044-45NL 4K
7489 16 x -
8101 256 x
8111 256 x
6599 16 x -
21L02/91L02 1024
74200 256 x
93421 256 x
MK4116(UP0416) 4K
x 1 Dynamic
4 Static
Static
I Static
4 Static
4 Static
l Static
x 1 Static
1 Static
1 Static
Dynamic 16 Pi
MK4027(UPD414) 16K Oyramic 16 Pin
AY-5-1013 30K Baud
ROMS
251312140) Char Gen -upper case
2513(3021) Char Gen -lower case
2516 Char Gen
MM5230 2048 BIT (512x4 on 256 x 8)
S 995
995
1095
1702A
5203 2048
82S23 32 x 8
82S123 32 x 8
74S287 1024
3601
Fames
Open C
Tnstate
Static
r~ -
395
Dynamic RAM
3 for 1 00|
Eprom 29 95
Eprom 59.95
Tn-State Bipolar 3 49
Open Collector Bipolar 2 95
SPECIAL REQUESTED ITEMS
FCM3817
AY -3-8500-1
AY-5-9100
AY -5 -9200
AY -5 -9500
AY-5-2376
9374
82S115
$5.00 HC90
7.50 4N33
1.50
HD0165
MCM6571
MCM6574
MCM6575
3.95
7.50
1.50
7.95
7205
ICM7045
ICM7207
ICM7208
ICM7209
MK50240
DS0026CH
TIL308
19.95 9368 3.95
24 95 LD110/111 25.00/set
7.50 95H90 11.95
19.95 MC3061P 3.50
7.50 MC4016 (74416) 7.50
17.50 MC1408L7 4.95
3.75 MC1408L8 5.75
10.50 74C922 9J5
parAtronics
Featured on February’s Front Cover of Popular Electronics
Logic Analyzer Kit Model iooa Model m
Model 100A
CLOCK CHIPS
MM5309 $9.95
MM5311 4 95
MM5312 4 95
MM5314 4 95
MM5316 6 95
MM5318 9 95
MM5369 2.95
MM5841 9 95
CT7[)01 5 9.5
$229. 00/kit
1 l 1 1 1
III ■
i««« ■
Analyzes any type of digital system
Checks data rates in excess of 8
million words per second
Trouble shoot TTL. CMOS. DTL. RTL.
Schottky and MOS families
Displays 16 logic states up to 8 digits wide
See ones and zeros displayed on your
CRT, octal or hexadecimal format
Tests circuits under actual operating conditions
• Easy to assemble — comes with step-by-step construction
manual which includes 80 pages on logic analyzer operation.
(Model 100A Manual - $4.95)
Some applications are
Troubleshooting microprocessor
address, instruction, and data flow
Examine contents of ROMS
Tracing operation of control logic
Checking counter and shift
register operation
Monitoring I/O sequences
Verifying proper system operations
during testing
PARATRONICS TRIGGER EXPANDER - Model 10
Adds 16 additional bits. Provides digital delay and qualification of input dock
and 24-bit trigger word. — Connects direct to Model 100A for integrated unit).
PRECISION
3 Vi-Digit Portable DMM
• Overload Protected
• 3' high LED Display
• Battery or AC operation
• Auto Zeroing
• Imv. iva. ( 0 1 onm resolution
• Overangs reading
• 10 meg input impendence
• DC Accuracy 1°» typical
Ranges: DC Voltage -0-1000V
AC Voltage 0-1000V
Freq Response 50-400 HZ
DC/AC Current: 0-1 00mA
Resistance 0-10 meg ohm
$9.00
20.00
Model 2800 Accessories:
$99.95 AC Adapter BC-28
Comes with test Rechargeable
leads, operating manual Batteries BP-26
and spare tuse Carrying Case LC-28 7.50
CONTINENTAL SPECIALTIES
PROTO BOARD 6 Other CS Prolo Boards
$15.95
(6" long X 4" wide)
Model 10 Kit - S229.00
Baseplate — $9.95
Model 10 Manual -$4 95
100 MHz 8-
• 20 Hz-inn MHz Range
• 6’ LED Display
• Crystal-controlled timebj
• Fully Automatic
> Portable — completely
X7.38*
Digit Counter
• Four power souces. e.
battenes. 110 or 220V with
* charger 12V with auto
lighter adapter and external
7.2-1 OV power supply
MAx-ioo $-|34.g5
ACCESSORIES FOR MAX 100:
Mobile Charger Eliminator
use power from car battery Model 180 — CLA $3.95
Charger/Eliminator
use 1 10 V AC Model 100 — CAi $9.95
PB100 -4.5 x 6
PB101 - 5.8" x 4.5"
PB102 - 7" x 4.5”
PB103 - 9" x 6"
PB104 - 9.5" x 8"
PB203 - 9.75 x 6Vi x 2%
PB203A - 9.75 x 6% x 2*4 129.95
(includes power supply)
$ 19.95
29.95
3995
59.95
79.95
80.00
LOGIC MONITOR
for DTL, HTL. TTL or CMOS Devices
$74.95
PROTO CLIPS
14 PIN
16 PIN
24 PIN
40 PIN
$4.50
4.75
8.50
13.75
DESIGN MATES
DM1 - Circuit Designer
$69.95
DM2 - Function Generator
$74.95
DM3 - RC Bridge
S74.95
a QT PROTO STRIPS
ItiiffiMgSiife • „
#ftoles
590
bus Strip
UT type
OT-59S
QT-59B
OT-47S
QT-47B
OT-35S
OT-35B
OT-18S
0T-I2S 120
QT-8S 80
0T-7S 70
Expen mentor 300
1250
250
10 00
2 29
850
200
4 75
3 75
3 25
300
$ 9.95
$1095
$5.00 Minimum Order — U.S. Funds Only
California Residents — Add 6% Salas Tax
Spec Sheets - 25c — Send 35c Stamp for 1978 Catalog
Dealer Information Available ^
*** 1978A
CATALOG
NOW
AVAILABLE
ELECTRONtCS
1021 -A HOWARD AVE., SAN CARLOS, CA. 94070
PHONE ORDERS WELCOME - (415) 592-8097
Advertised Prices Good Thru May
J1
The Incredible
Penny whistle 103’ 1
$129.95 K it Only
The Pennywhislle 103 is capable of recording data to and from audio tape without
critical speed requirements lor the recorder and it is able to communicate directly
with another modem and terminal for telephone "hamming ' and communications
for the deaf In addibon. it is free of critical adjustments and is built with non-precision,
readily available parts
Data Transmission Method . . .Frequency-Shift Keyeig. full-duplex (half-duplex
selectable).
Maximum Oata Rate 300 Baud
Data Formal Asynchronous Senal (return to mark level required
between each character).
Receive Channel Frequencies . . .2025 Hz for space: 2225 Hz for mark.
Transmit Channel Frequencies ..Switch selectable: Low (normal) = 1070 space.
1270 nark; High = 025 space. 2225 mark.
Receive Sensitivity -46 dbm accoustically coupled
„ Transmit Level -15 dbm nominal. Adjustable from -6 dbm
to 20 dbm
Receive Frequency Tolerance . . .Frequency reference automatically adjusts to
allow lor operation between 1800 Hz and 2400 Hz
Digital Oata Interlace EIA RS-232C or 20 mA current loop Irecefver is
optoisolated and non-polar)
Power Requirements 120 VAC, single phase. 10 Watts
Physical All components mount on a single 5' by 9"
printed circuit board. Art components included
Requires a VOM. Audio Oscillator, Frequency Cornier aid or Oscilloscope to ahgi
the 3
rd
Hand
$9.95 each
•Leaves two hands free for
working
* Clamps on edge of bench, table
or work bench
* Position board on angle or flat
position for soldering or clipping
* Sturdy, aluminum construction
for hobbyist, manufacturer or
school rooms
DIGITAL STOPWATCH
• Bright 6 Digit LED Display
• Times to 59 minutes 59 59 seconds
• Crystal Controlled Time Base
• Three Stopwatches m One
Times Srngle Event — Split & Taylor
• Size 4 5 x 2 15 x 90 (44 ounces)
• Uses 3 Penute Cells
Kit — $39.95
Assembled — $49.95
Heavy Duty Cany Case $5.95
Stop Watch Chip Only (7205) $19.95
3V2DIGIT DPM KIT
New Bipolar Unit
• Auto Zeroing
• .5" LED
Model KB500 DPM Kit
Model KB503 5V Power Kit
Auto Polarity
Low Power
Single 1C Unit
$49.00
$17.50
JE700 CLOCK
The JE 700 is a tow cost digital clock, b
is a very high quality unit Trie umi fea-
tures a simulated walnui case wnh di-
mensions of 6 x 2'z x 1 it utilizes a
MAN 72 high brightness readout and the
MM5314 dock chip
KIT ONLY
$16.95
HEXADECIMAL
ENCODER 19-KEY PAD
.1-0
• ABCDEF
• Shift Key
• 2 Optional Keys
$10.95 each
New 63 KEY KEYBOARD
$29.95
IN STOCK
-RCTTm rTTode? Chip (encodes 16 Keys)
AY-6-2376 Encoder Chio (encodes 88 Keys)
This keyboard features 63
coded SPST keys unattached to
any kind of P C B A very solid
molded plastic 13 x 4 base
suits most applications
$7.96 M.
$14.95 M
JE803 PROBE —
The Logic Probe is a umi which is l$r the most pan — " ^
mdespensibte m trouble shooting logic families / _____ m *
TTL DTL RTL CMOS it derives the power it ^ *
needs to operate directly oft of the circuit undei ” — —
lest drawing a scant 10 mA max II uses a MAN3
readout lo indicate any ot me following stales by
mese symbols (Hi 1 1 LOW) - o (PULSE) P The n ....
Probe can delect high ireguency pulses to 45 MHz JbS.SO iBT Kit
h can t be used ai MOS levels or circuit damage
“ ffsun printed circuit board
T 2 L 5V 1A Supply
This is a standaro TTL power supply using the well known
LM309K regulatoi !C to provide a solid 1 AMP of current at 5
volts We try to make things easy tor you by providing
eveiythmg you need m one package including the hardware j
$9.95 Per Kij/
tor only
JE225
133
OVENAIRE ULTRA PRECISION CRYSTAL OSCILLATOR
Your computer is only as good as its clock. We have been fortunate in
acquiring a lot of OVENAIRE precision crystal oscillators. Model OSC
67-11-A-3. The output frequency of these oscillators is 3.840 Mhz.
This frequency readily divides into many useable frequencies with
the use of standard SN7400 series ICs. Among the many frequencies
are 640 Khz, 60 Khz, 32 Khz, 20 Khz, 10 Khz, 6 Khz, 1 Khz, 600 Hz, 100 Hz, 60 Hz, 50 Hz. and many
more. We provide data showing the ICs needed to get these frequencies. The oscillator is precise to 2 parts
per million, and is adjustable to even greater precision. Ideal for computers, frequency standars, clocks etc.
This oscillator is a current production item, and the one piece price at the factory is $134.50. In lots of
100 the price is $49.80, so our price of $14.95 each is a fantastic bargain. 1 5/8"x 2"x 5/8". PC mount.
Voltages required are 5 VDC and 12 VDC. Output is TTL compatible 5 VDC. Sketch at left shows the
complete unit, and an inside view.
STOCK N0.5592K Ovenaire Precision Crystal Oscillator $14.95 ea. 2/28.00
NEW POWER TRANSFORMERS
Tapped 115 V primary. Secondary either 12.5 or 14volts @2 A. 3 1/8"x3"x2%". 3 lbs.
STOCK NO.9031 K $3.95 ea. 2/6.00
Primary 115V. Sec.1. 16.5V@ 1.5A. Sec.2 16V@ 3.5A. Sec.3 9.5V@ 3.5 A. Sec.4, 130 V@2A. lOLbs.
STOCK N0.6677K $10.95 ea. 2/20.00
SQUIRREL CAGE BLOWERS
Keep your valuable equipment from overheating. We have squirrel cage blowers,
made by REDMOND, 115V, 60 Hz. .78 A. 3000 RPM. Removed from equipment.
STOCK N0.9325K 4 $9.95 ea. 2/18.00
WIRE WRAP BOARDS LOADED WITH 7400 SERIES ICs
Since last summer, we have been selling 2 wire wrap boards,
Our Stock No. 6558K with approximately 100 sockets, and
our Stock No. 6559K with approximately 45 sockets.
These have been successful, based on your orders and
reorders. We now have the same boards, but with the
sockets still containing the original SN7400 series ICs that
were used in the computer that these boards were designed
for. We checked the value of these ICs, against the lowest
price ICs in several Electronics magazine, and found that
at the lowest possible surplus prices, the values of the ICs on the 100 socket board ran to over $40.00. A sample of
some of the chips on the board we looked at are as follows: 74H87, 7486, 74107, 7451, 7400, 7404, 7495, 7493,
7492, 74193, 7489 and many others, to numerous to mention. Also on some boards, are a few linears, and phase
locked loops. Not everyone needs every chip, but if you are working at all with TTL, this is a great opportunity to get
an inventory of the most useful chips at a ridiculous price. We are selling the 100 socket board with about 100 chips,
for $10.00 more than the board itself, and the 45 socket chip for $5.00 more than the board itself. We will also include
with each board, 2 edge connectors with the 100 socket board, and 1 edge connector with the 45 socket board.
Connectors NYLON CONNECTORS -
Per Pkg.
Type No.
Class
Description
Eo. Pkg
5
1625-IPRT
Min. (.062')
1 Circuit
SI .75
3
I625-2PRT
»
2 Circuit
1.90
3
I625-3PRT
»
3 Circuit
2.10
2
I625-4PRT
■
4 Circuit
2.10
2
I625-5PRT
5 Circuit
2.20
2
I625-6PRT
»
6 Circuit
2.35
l
I649-8PRT
»
8 Circuit
1.55
1
I625-9PRT
9 Circuit
1.75
1
I625-I2PRT
«
12 Circuit
1.90
1
I625-I5PRT
«
15 Circuit
2.30
1
I625-24PRT
■
24 Circuit
3.25
1
I772-36PRT
■
36 Circuit
4.55
5
I6I9PRT
Std.(.093")
1 Circuit
1.75
3
I545PRT
«
2 Circuit
1.90
3
I396PRT
u
3 Circuit
2.10
2
I490PRT
•1
4 Circuit
2.10
2
I653PRT
II
5 Circuit
2.20
2
1261 PRT
«
6 Circuit
2.35
1
I292PRT
«
9 Circuit
1.80
1
I360PRT
»
12 Circuit
1.90
1
I375PRT
-
15 Circuit
2.45
Prototype hand tools combine efficiency with
hmited production runs
HT 1919 lor 093“ pm <to term.nats
HT 1921 lor 062" pm oa term.
Econo-Extractor removes terminal from nylon connector housing with
smoothness and ease
HT-2054 for extracting 093“ pin d>a terminals
HT-2023 for extracting 062' pm dia terminals
Deluxe ejector tools, spring loaded for simple efficient removal o
Irpm nylon connector housing extracts either male or lemaie terminal
pm diameter
HT 2038 for extracting 093“ pm dia terminals
HTt 01 0-282 Replacement tip for HT 2038
HT 2285 for extracting 062“ pm dia terminals
HT-1672 3 replacement tip for HT 2285
INTEGRATED CIRCUITS
555 Timer 8 pin mini-DIP
741 Compensated OP-Amp 8 pin DIP
LM 1889N RF Video Modulator
CA3130 Bipolar/Mos-FET Op Amp
CA3140 MOS-FET Op Amp, Bi-polar out
LM3909 Lo Voltage Led Pulser
LM391 1 Temp Control CHIP
Signetics 2504TA 1024 bit S.R. memory (I404A).
MCM 6571P Character Generator
MCM6571AP Character Generator
LM399H Temp Stabilized Zener
AF100-1CJ Active Filter, State Variable
LM2907N Tachometer F/V Converter
LM1812N Ultra sonic Transceiver
LM 181 5 Adaptive Sense Amp for Tachometer
S2518B Hex 32 bit Shift Register $2.95
TLI70 TO-92 Hall effect switch w/spec sheets 1 .25
MC14409P Telephone Rotary Pulser 10.98
MC14419P Touch Pad Converter for 14409 4.25
MCI 441 IP Baud Rate Generator 11.98
MC14412VP CMOS Modem Chip 16.95
MM57109N Number Cruncher Micro 18.95
74C915 7 Segment to BCD Converter 2.99
74C922 16 Key Keyboard Encoder 6.35
74C923 20 key Keyboard Encoder 6.45
74C925 4 Decade Counter w/latches 1 2.00
74C926 4 Decade Counter w/carry 12.00
74C935-1 3% Digit DVM CMOS Chip 16.98
9601 Retriggerable One shot 50
MC40I5P Hi Speed quad "D" low power TTL $1.00
DATA ACQUISITIONS SUBSYSTEM
ADC0817 is a 40 pin CMOS I.C. with an 8-bit analog-
to-digital converter, a 16 channel multiplexer and micro
processor compatible control logic. Converter features
high impedance chopper stabilized comparator, voltage
divider with analog switch tree and a sucessive approx-
imation register. Latched Tri-State outputs for easy
u-processor interface* Require only 15 mW of power
from single 5V supply. Fast, IOOuS conversion time.
ADC0817 (40 pin DIP) $29.88
Spec sheets $1.00
Weal for prototype or
$8 95 each
S 8 95 each
TO-220 Mounting Kit
6I06PB
I
60I3B&L_^
I 6052B
« ?
Economical 1 piece heat sinks for plastic power parts in TO-220
and Motorola cases 77, 90, 199, and TO-126. All are black
anodized aluminum. "B" series is anodized after forming. "PB"
series is anodized prior to forming.
THM 6030PB Vertical
THM 6045B Slip Over
THM 6070 B
THM 6071 B
THM 6I06PB
THM 6I07PB
THM 60I3B
Extra Disipation Horiz
Top Hat for 6070
Flat With Fingers
Smaller Size Flat
TO-3 Diamond
THM 6024- U Unfinished TO-92
THM-6052B TO-3 Square
TO-220 6 Piece Mounting Kit -
25<?, 5/$1.00,
306, 4/$l .00,
.356, 3/$l .00,
356, 3/$l .00,
306, 4/$l .00,
256, 5/$1.00,
696, 4/$2.50,
10 /$ 1 . 00 ,
606, 5/$2.50,
Handy Package.
10/$1 .90
10/$2.00
10/$3.00
10/$3.00
l 0/$2.00
I0/$1 .90
10/$5.00
100/55.00
10/54.25
..256
P. C. BOARD TERMINAL STRIP
Molded body encloses postive screw activated clamp
which will accomodate wire sizes 14-30 AWG. Contacts
and pins are solder plated copper. Pins are on .200
inch (5.08mM) for standard P. C. mounting. lOAmp
rating. Compare our prices before you buy
4 pole TS-2504
8 pole TS-2508
12 pole TS— 251 2
S-100 BUS CONNECTORS (IMSAI TYPE)
Gold, Solder tail for Mother boards $4,50 / 4/$l7.00
Tin-Nickel, (NASGLO) Solder toil $3.75,4/$l4.00
Gold, wire-wrap $4.50,4/517.00
Tin-Nickel, (NASGLO) wire-wrap $3.75,4/514.00
HEX DARLINGTON ARRAY
MCI4I3P is a 16 pin DIP package with (6) 50V 500mA
Darlington pairs.
MCI4I3P $1.59
Specs/Apps 306
I.C. SOCKETS
Lo Profile Tin Solder Tail Dip Sockets
10/Si .50 100/S14.00 1000/51 20.00
10/S1.70 100/S16.00 1000/S140.00
10/$ 1.90 100/S18.00 1000/S1 60.00
VOLTAGE REGULATORS
7805-06 08- 12-1 5-24 T0220 95*
7905-06-08 1 2- 1 5-24 TO 220 95«
70LO5A-12-15 4% 100 mA TO-92 Plastic
78H05KC 5V 5A TO-3
78H12KC 12V 5ATO-3
78H15KC 15V 5ATO-3
Lm3 1 7 K 1 . 5 A Adjustable TO-3
Lm317T ] .5A Adjustable TO-220
Lm317MP ,5A Adjustable T 0-202
TL430C Adjustable Zener-Think About It
TL497C Switching Reg. & Inductor
RCA CA 3085 100 mA Adjustable
ADJUSTABLE NEGATIVE REGULATOR
LM337 ?s the compliment to the popular LM3I7
adjustable regulator. Capable of 1 .5Amp from
to - 37V.
LM337K (TO-3 Metal)
LM337T (TO-220 Plastic)
Specs and applications
5/S4.50
5/S4.50
50c
9.15
9.15
9.15
4.99
3.99
13.95
1.50
9.50
.60
positive
- 1.2V
. $5.99
. $4.65
606
DIODES AND BRIDGES
IN4003 200V lamp 12/$1.00
IN4004 400 V 1 amp 10/S1.00
I N4 148 Hi Speed Signal 15/S1.00 100/S5.00
D 600 1 15 V, 100 mA Hi Speed Signal 20/$1.00
D2131 200 V, 25A Stud 85c
D2 1 35 400 V, 25A Stud 1 .00
D2138 600 V, 25A Stud 1.55
D3289R 200 V, 160A Stud Anode 5.85
D3909-4 50 V, 45A Fast Recovery 2.00
IN4732A-47A 1W 5% Zeners 4/$1.00
1 3 Assorted Brand New Zener Diodes 1 .00
50V 3 amp Epoxy Bridge 79c
200V 30 amp Bridge 2.00
600V 4 amp Epoxy Bridge 1 .49
600V 3 amp Stud Bridoe „ , 89
SI -2 200V, 1.5A Gold Leads 15/51.00
D1A-0030 30V DIAC 10/S1.00
100V 2.5 WATT ZENER
IN5051A Epoxy Zener with heavy silver leads 256
MISCELLANEOUS
RG-1 74 Miniature 50 T2 coax 5074.25
WSU-30 Wire Wrap/unwrap tool 5.95
WSU-30M Modified Wrap/unwrap tool 6.95
BW-630 Battery Operated Wrap Tool 34.95
-Free Wire with any Wrap Tool —
Miniature Square .05/1 00V Monolithic Cap 10/2.00
FND5002 .125" C.C. 7 Seg. Read-out. .496, 10/53.95
2N4036 90V, I A PNP Silicon TO-5 .50
2N6I0I 80V, IQA NON HI GAIN TO-220 .50
6.3 VCT, 1 .2A Transformer F4 1 X 2.49
12V, 1A Transformer with 6' Power Cord 2.88
Fairchild Linear Data Book. Huge volume has 17 chap-
ters of data, applications and definitions. Probably
the biggest volume of its type today $4.50ppd
Fairchild Bipolar Memory Data Book. ROMS, PROMS
and RAMS are covered in this work with full engineer-
ing data. Price includes shipping ! $3.75
NEW NATIONAL DATA BOOK
1978 Edition of Nationals discrete devices data book.
9 big sections of data plus extensive glossary of terms
and test diagrams. Covers bi-polar small signal and
power devices as well as F.E.T.s $3.95
Raytheon Micro Computer Components Book. Covers
Raytheons 1 offerings in the field from the micro-puter
to memories and adapters $2.95ppd
w
-tek. inc.
tRI
7808 North 27th Avenue
Phoenix. Arizona 85021
(602) 005 0352
Please 9 ,V« street address for UPS shipping when possiDle.
C.O.O. NO parcel post C.O.O.
UPS C.O.D. Add 85« to order,
correspondence i
please use separate u
Orders less than *10 (SIS foreign) please add SI
handling.
Prices are subject to change without notice.
Any refunds will be by check, no! credit vouchers.
and 'furnished with’
___ cannot ship in 30
notified of the expected shipping date
postage paid card v
T1
we pay surtaca shipping <
it class, special handling, etc.)
Charge card telephone orders (520 min.) t
accepted 9-5:30 P.M. except weekends.
Telephone 995-9352. No collect calls pleat
135
DIODES/ZENERS
1N914
lOOv
10mA
.05
8-pin
pcb
.25
ww
.45
1N4005
600v
1 A
.08
14-pin
pcb
.25
ww
.40
1N4007
lOOOv
1 A
.15
16-pin
pcb
.25
ww
.40
1N4148
75v
10mA
.05
18-pin
pcb
.25
ww
.75
1N753A
6.2v
z
.25
22-pin
pcb
.45
ww
1.25
1N758A
lOv
z
.25
24-pin
pcb
.35
ww
1.10
1N759A
12v
z
.25
28-pin
pcb
.35
ww
1.45
1N4733
5.1v
z
.25
40-pin
pcb
.50
ww
1.25
1N5243
13v
z
.25
Molex pins .01
To-3 Sockets
.45
1N5244B
1N5245B
14v
15v
z
z
.25
.25
2 Amp Bridge
100-prv
1.20
25 Amp Bridge
200-prv
1.95
SOCKETS/BRIDGES
TRANSISTORS, LEDS, etc.
2N2222A NPN (2N2222 Plastic .10)
2N2907A
2N3906
2 N 3904
2N3054
2N3055
TIPI 25
PNP
PNP
NPN
NPN
NPN
PNP
(Plastic)
(Plastic)
15A 60v
Darlington
LED Green, Red, Clear, Yellow
D.L.747 7 seg 5/8“ High com-anode
XAN72 7 seg com-anode (Red)
MAN71 7 seg com-anode (Red)
MAN361 0 7 seg com-anode (Orange)
MAN82A 7 seg com-anode (Yellow)
MAN74A 7 seg corn-cathode (Red)
FND359 7 seg corn-cathode (Red)
.15
.15
.10
.10
.35
.50
.35
.15
1.95
1.25
1.25
1.25
1.25
1.50
1.25
/€ MOS
- T T
L -
4000
.15
7400
.15
7473
.25
74176
1.25
74H72
.45
74S133
.40
4001
.15
7401
.15
7474
.30
74180
.75
74H101
.75
74S140
.55
4002
.20
7402
.20
7475
.35
74181
2.25
74H103
.75
74S151
.30
4004
3.95
7403
.20
7476
.40
74182
.95
74H106
.95
74S153
.35
4006
.95
7404
.15
7480
.55
74190
1.75
74S157
.75
4007
.35
7405
.25
7481
.75
74191
1.05
74 LOO
.25
74S158
.30
4008
.95
7406
.35
7483
.95
74192
.75
74L02
.25
74S194
1.05
4009
.45
7407
.55
7485
.75
74193
.85
74L03
.30
74S257 (8123)
1.05
4010
.45
7408
.25
7486
.25
74194
1.25
74L04
.30
4011
.20
7409
.15
7489
1.35
74195
.95
74L10
.30
74LS00
.25
4012
.20
7410 .
.10
7490
.55
74196
1.25
74L20
.35
74LS01
.35
4013
.40
7411
.25
7491
.95
74197
1.25
74L30
.45
74LS02
.35
4014
.95
7412
.30
7492
.95
74198
2.35
74L47
1.95
74LS04
.30
4015
.90
7413
:35
7493
.35 ,
74221
1.00
74L51
.45
74LS05
.45
4016
.35
7414
1.10
7494
.75
74367
.85
74L55
.65
74LS08
.25
4017
1.10
7416
.25
7495
.60 "
74L72
.45
74LS09
.35
4018
1.10
7417
.40
7496
.80
751 08A
.35
74L73
.40
74LS10
.35
4019
.50
7420
.15
74100
1.15
75110
.35
74L74
.45
74LS11
.35
4020
.85
7426
.30
74107
.35
75491
.50
74L75
.55
74LS20
.25
4021
1.00
7427
.45
74121
.35
75492
.50
74L93
.55
74LS21
.25
4022
.85
7430
.15
74122
.55
74L123
.85
74LS22
.25
4023
.25
7432
.30
74123
.55
74H00
.15
74LS32
.40
4024
.75
7437
.30
74125
.45
74H01
.25
74S00
.35
74LS37
.35
4025
.30
7438
.35
74126
.35
74H04
.20
74S02
.35
74LS40
.45
4026
1.95
7440 *
.25
74132
1.35
74H05
.20
74S03
.30
74LS42
1.10
4027
.50
7441
1.15
74141
.90
74H08
.35
74S04
.30
74LS51
.50
4028
.95
7442
.45
74150
.85
74H10
235
74S05
.35
74LS74
.65
4030
.35
7443
.65
74151
.65
74H11
.35
74S08
.35
74LS86
.65
4033
1.50
7444
.45
74153
.75
74H15
.45
74S10
.35
74LS90
.95
4034
2.45
7445
.65
74154
.95
74H20
.30
74S11
.35
74LS93
.95
4035
1.25
7446
.95
74156
.95
74H21
.25
74S20
.35
74 LSI 07
.85
4040
1.35
7447
.95
74157
.65
74H22
.40
74S40
.20
74 LSI 23
1.00
4041
.69
7448
.65
74161
.85
74H30
.20
74S50
.20
74LS151
.95
4042
.95
7450
.25
74163
.85
74H40
.25
74S51
.25
74 LSI 53
1.20
4043
.95
7451
.25
74164
.60
74H50
.25
74S64
.20
74 LSI 57
.85
4044
.95
7453
.20
74165
1.50
74H51
.25
74S74
.35
74 LSI 64
1.90
4046
1.75
7454
.25
74166
1.35
74H52
.15
74S112
.60
74LS367
.75
4049
.45
7460
.40
74175
.80
74H53J
.25
74S114
.65
74LS368
.75
4050
.45
7470
.45
74H55
.20
74C04
.25
4066
.95
7472
.40
74C151
2.25
4069
4071
4081
4082
MCI 4409
MCI 441 9
.40
.35
.70
.45
14.50
4.85
MCT2
.95
LINEARS, REGULATORS, etc.
9000 SERIES
9301 .85
95H03 1.10
9309 .35
9601 .45
9322 .75
9602 .45
MICRO'S,
RAMS,
CPU'S,
ETC.
74S188
3.00
1702 A
4.50
MM5314
3.00
MM5316
3.50
2102-1
1.45
2102L-1
1.75
TR1602B
4.50
TMS 4044-45NL 14.50
8080 AD
12.00
8T13
1.50
8T23
1.50
8T24
2.00*
8T97
1.00
2107B-4, A
4.00
2708
11.50
8038
3.95
LM320T5
1.65
LM340K1 5
1.25
LM723
.50
LM201
.75
LM320T12
1.65
LM340K18
1.25
LM725N
2.50
LM301
.45
LM320T15
1.65
LM340K24
.95
LM739
1.50
LM308 (Mini)
.95
LM324N
.95
78L05
.75
LM741 (8-14) .25
LM309H
.65
LM339
.95
78L12
.75
LM747
1.10
LM309K (34qk-5*85
7805 (340T5)
.95
78L15
.75
LM1307
1.25
LM310
1.15
LM340T12
1.00
78M05
.75
LM1458
.95
LM31 ID (Mini)
.75
LM340T 1 5
1.00
LM373
2.95
LM3900
.50
LM318 (Mini)
.95
LM340T18
1.00
LM380(8-i4PiN).95
LM75451
.65
LM320K5(7905)1 .65
LM340T24
.95
LM709 (8,14 pin).25
NE555
.50
LM320K1 2
1.65
LM340K12
1.65
LM711
.45
NE556
.95
NE565
.95
INTEGRATED CIRCUITS UNLIMITED „
NE566
NE567
1.75
1.35
7889 Clairemont Mesa Boulevard, San Diego, California 92111
(714) 278-4394 (Calif. Res.)
All orders shipped prepaid No minimum
Open accounts invited COD orders accepted
SPECIAL
DISCOUNTS
Total Order Deduct
$35 - $99 5%
$100 -$300 10%
$301 -$1000 15%
$1000 -Up 20%
Discounts available at OEM Quantities California Residents add 6% Sales Tax
All IC's Prime/Guaranteed. All orders shipped same day received.
24 Hour Toll Free Phone 1 -800- 854-221 1 American Express / BankAmericard / Visa / MasterCharge
THE CHIPS ARE DOWN !
8K NOW JUST $149 ASSEMBLED
George Morrow- Inventor 8- Designer
1201 10th Street Berkeley, CA 94710
*ECONORAM is a trademark of Godbout Electronics
Thinker Toys™ brings down the
high cost of adding big memory
capacity to your S-100 system!
The ECONORAM III* 8Kx8 (by
Morrow's Micro- Stuff) comes fully
assembled, burned in, tested and
fully warranted for one full year —
for just $149!
It's configured as two individually
addressable 4K blocks. And it typically
consumes less than one-half the
power of any competitively-priced
memory.
Obviously, our new ECONORAM
III* 8Kx8 isn't just another cheap kit.
It's a design breakthrough in dynamic
memory that gives you guaranteed
reliability with tremendous savings
in cost and power.
You save on cost and power
because ECONORAM III* is the first
fully reliable dynamic memory. It's
unique SynchroFresh™ refresh cir-
cuitry weaves itself into the natural
timing of the S-100 bus to provide
simple, invisible and absolutely
reliable refresh . . . with fewer parts,
less power, lower cost and no
interference with computer speed
or timing.
You just plug ECONORAM III* into your S-100 bus and run perfectly
for at least one full year. Or your money fully refunded. It's that simple.
So don't let the high cost of memory keep your system down. ECONO-
RAM IH*8Kx8 memory is available now at leading computer shops. Or ask
your nearest computer shop to order it for you.
Or, if unavailable locally, write direct to Thinker Toys™ for specs and
ordering. BAC/MC orders can be placed by phone to 415-527-7548
(10-4 PST). Add $3 for handling; Cal. res. add tax.
Morrow’s Micro-Stuff for
nker Toys
137
HARD COPY STORAGE A PROBLEM?,
Kilobaud, as thick as it is, is more like a floppy when it
comes to standing on the bookshelf. Try the Kilobaud
Library Shelf Boxes, . . . sturdy corrugated white dirt re
sistant cardboard boxes which will keep them from flop
ping around. We have self-sticking labels for the boxes
too, not only for Kilobaud, but also for 73 Magazine . .
and for Personal Computing, Radio Electronics, Inter-
face Age, and Byte . Ask for whatever stickers you want
with your box order. Hams may want out labels for CQ,
QST or Ham Radio. They hold a full year of Kilo-
baud ... or 73. Your magazine library is your prime
reference, keep it handy and keep it neat with these
strong library shelf boxes . . . $2.00 for the first box and
$1.50 for each additional box. Be sure to specify which
labels we should send. Have your credit card handy and
call our toll-free order number 800-258-5473, or use the
o der card in the back of the magazine and mail to
KILOBAUD LIBRARY SHELF BOXES Peterborough, NH 03458
IT’S EASY AND FUN
TO BUILD YOUR OWN TEST EQUIPMENT
WITH ICs
. TO! r
IC TEST EQUIPMENT S4.85*
ICs have greatly simplified even the most sophisticated pieces of test
equipment, making them fun to build. You can save thousands of dollars
by building your own equipment and have a test lab which would make a
university jealous.
A brand new book, IC TEST EQUIPMENT, has construction projects
for making 37 pieces of test equipment. Square wave generator, pulse
generator, timer, audio sythesizer, AF8K generator, sync generator,
counters, capacity meter, etc.
Order this book today and get started building your own lab.
Use the order card in the back of the magazine or itemize your order on
a separate piece of paper and mail to: KILOBAUD BOOK NOOK, Peter-
borough NH 03458 or phone toll free 800-258-5473
Be sure to include check or detailed credit information. *Add 81 ship-
ping and handling charge for each order.
73
I changing
a lots of ideas
l(D
Every month there arc computer articles in 73 ... a it of them. Fact is, since February 1976, 73 Magazine
has published articles directed to the Computerist and Soon-to-be Computerist. There are also a lot of articles
that computer hobbyists will be needing to read which are not exactly computer articles such as on regulated
power supplies ... on making printed circuit boards ... on how various circuits work . . . things like that which
hardware men in particular need to read . . . and which software people need even more, since they are a bit
behind on hardware.
In recent issues there have been articles on computerized satellite tracking (with software), RTTY using a uP,
using old (inexpensive) Teletypes, building a Polymorphic video board, making instant PC boards using the new
color-key technique, the TTL one-shot, what computers can and can't do, a hamshack file handler (software), the
bit explosion — 8-12-16?, backward branch the easy way with the 6800, the hexadecimal . . . etc.
Any one of these articles could easily be worth the cost of a full year of 73. One good program could save youd
days of work. One good interface project could make an enormous difference. In general, 73 tries to present not!
too complicated construction projects . . . things you can make in a day or two.
Yes! Enter my subscription to 73 MAGAZINE for 1 year starting with the next published issue - only $15.00.1
Name Call
State .
. Zip
bout
Computers
7
S enclosed □ Cash □ Check □ Money Order
Bill: □ Master Charge □ Bank Americard/ VISA □ American Express
Expiration date _
5/78
One of the fundamental policies is that no articles will be published in both 73 and Kilobaud.
This is, in a way, unfair because it keeps some great computer articles away from computerists.
You really must get both magazines to keep up to date with what is going on. When you
subscribe to both, you will not be getting duplication.
Signature
□ Bill me direct (I've signed above) Allow 6 weeks for subscription processing.
Toll Free Subscription Number: (800) 258-5473
This offer expires in 60 days. 73 MAGAZINE. PETERBOROUGH NH 03458
138
Herb Waite looks up from
behind his copy of KB.
Who’s Behind
the Kilobaud?
In between selling sub-
scriptions and seeing the
other exhibits at computer
shows, Wayne snaps pic-
tures of people you have
been seeing and will be see-
ing at shows. How many of
them can you recognize
behind the Kilobauds?
Still waiting for the
answer from last month’s
clues? Drop by the PerSci
booth at the next computer
show and talk with Herb
. . . he’ll tell you about one
of the most popular disk
drives available today as
well as confirm that KILO-
BAUD is a favorite maga-
zine of his . . . and
thousands of other com-
puterists.
Kilobaud is the most
read magazine in the
microcomputer field. It
hasn’t been difficult to find
key industry people reading
KB!
kilobaud
READ IT YOURSELF!
If you are not yet a sub-
scriber to Kilobaud, we
need you and you need us.
There is a lot of data in Kilo-
baud that you will want to
have on hand . . . the
magazine is like a continu-
ing encyclopedia of micro-
computing and programs.
You never know when you
are desperately going to
need something from a past
issue . . . and you’ll want it
immediately. The cost per
year is not significant . . .
$15 at present for $24 worth
of magazines.
You may have noticed
that Kilobaud has more arti-
cles than any of the other
magazines. If you’ll keep
track of how long it takes
you to read Kilobaud as
compared to the other com-
puter magazines, you’ll see
you are getting a great
bargain at $15 per year.
WHO’S BEHIND
THE
KILOBAUD?
-j
M o
1
I®
1
IL
2
: S
SSi
Who’s This?
In addition to being a
frequent author in KILO-
BAUD, the chap behind the
KILOBAUD is also the
owner of a microcomputer
store. As a further hint, the
store is in a small city of
about 40,000, even so, busi-
ness is growing nicely and
the store is thriving. Not
bad when you consider
there are two other com-
puter stores in the same
town . . . and a fourth in
the works!
SUBSCRIBE TODAY!
For Instant Subscriptions Call Our Toll Free Number
l-(800) 258-5473
Have your credit card handy!
Kilobaud • Peterborough NH 03458
139
& s H % * ? % 1 # • ■
•THE STORY OF COMPUTERS by Donald
D. Spencer is to computer books what Dick
and Jane is to novels . . . extremely elemen-
tary, gives the non-computerist a fair idea of
what the hobbyist is talking about when he
speaks computer lingo. Attempts to explain
what computers are and can do to a spouse,
child or any un-electronics-minded friend.
$4.95.*
• HOBBY COMPUTERS ARE HERE If you
(or a friend) want to come up to speed on
how computers work . . . hardware and soft-
ware . . . this is an excellent book. It starts
with the fundamentals and Explains the
circuits, the basics of programming, along
with a couple of TVT construction projects,
ASCII-Baudot, etc. This book has the highest
recommendations as a teaching aid for new-
comers. $4.95.*
• THE NEW HOBBY COMPUTERS! This
book takes it from where "Hobby Computers
Are Here" leaves off, with chapters on Large
Scale Integration, how to choose a micro-
processor chip, an introduction to program-
ming, low cost I/O for a computer, computer
arithmetic, checking memory boards, a
Baudot monitor/editor system, an audible
logic probe for finding those tough problems,
a ham's computer, a computer QSO machine
... and much, much more! $4.95.*
• HOME COMPUTERS: 2 10 Questions &
Answers by Rich Didday. Two books aimed
exclusively at the novice computer hobbyist/
home computer user. Written in a rather
unusual style which has a beginner asking
questions which are answered by a person
with a substantial background in computers
and personal computing. The questions are
just the kind beginners come up with . . . and
the answers are presented in easy-to-under-
stand terms (usually with a diagram to
illustrate the point). Both the hardware and
software aspects of home computing are
covered from A to Z. An index in both books
makes them ideal as reference material for
anyone . Volume I: Hardware — $7.95*;
Volume 2: Software — $6.95*.
•MICROCOMPUTER PRIMER by Mitchell
Waite and Michael Pardee. Describes basic
computer theory, explains numbering sys-
tems, and introduces the reader to the world
of programming. Describes the world of
microcomputing in "real world" terminology.
No better way of getting involved with the
exciting new hobby of microcomputing.
$7.95.*
beginner’s •introductory
wmtmm
; |
• WHAT TO DO AFTER YOU HIT RETURN
PCC's first book of computer games ... 48
different computer games you can play in
BASIC ... programs, descriptions, muchly
illustrated. Lunar landing, Hammurabi, King,
Civel 2, Qubic 5, Taxman, Star Trek, Crash,
Market, etc. $8.00.*
• SCELBI'S GALAXY GAME FOR THE
"6800" Here's a new twist in computer games
by Robert Findley/Raymond Edwards.
"Galaxy" pits the operator of a spaceship
against alien craft, as well as such variables as
speed, time, and ammunition. No two games
are the same! $14.95.*
• 101 BASIC COMPUTER GAMES Okay, so
once you get your computer up and running
L
computer games
AT TO ry-v
* y OUHff
ETURN
Book of
ier Gomes
• MICROPROCESSORS FROM CHIPS TO
SYSTEMS by Rodnay Zaks is a complete and
detailed introduction to microprocessors and
microcomputer systems. No preliminary
knowledge of computers or microprocessors is
required to read this book, although a basic
engineering knowledge is naturally an
advantage. Intended for all wishing to under-
stand the concepts, techniques and com-
ponents of microprocessors in a short time.
$9.95.*
• INTRODUCTION TO MICROPRO-
CESSORS by Charles Rockwell of MICRO-
LOG is an ideal reference for the individual
desiring to understand the hardware aspects
of microprocessor systems. Describes the
hardware details of computer devices in terms
the beginner can understand, instead of treat-
ing the micro chip as a "black box." General
information about hardware systems is pro-
vided. Specific systems are not described and
programming is only briefly discussed. $17.50
US and Canada, $20 elsewhere.*
• AN INTRODUCTION TO MICROCOM-
PUTERS, VOLS. 1 AND 2 by Adam Osborne
Associates, are references dealing with micro-
computer architecture in general and specifi-
cally with details about most of the common
chips. These books are not software-oriented,
but are invaluable for the hobbyist who is
into building his own interfaces and pro-
cessors. Volume 1 is dedicated to general
hardware theory related to micros, and
Volume 2 discusses the practical details of
each micro chip. (Detailed review in Kilobaud
#2) Volume 1 - $7.50*; Volume II -
$12.50.*
• HOME COMPUTERS: A BEGNINNERS
GLOSSARY AND GUIDE this book is in-
tended as a quick reference source for begin-
ners. Included is a general introduction to
microcomputers, a simple application &
sample system, the history of microcomputers
& their uses, and an introduction to same
actual equipment. A chapter on number
systems includes a number conversion chart,
binary arithmetic from conversions to divi-
sions, and a discussion of octal and hexa-
decimal numbers. A good background to read
technical literature and computer equipment
specifications. $6.95.*
in BASIC, then what? Then you need some
programs in BASIC, that's what. This book
has 101 games for you, from very simple to
real buggers. You get the games, a description
of the games, the listing to put in your
computer and a sample run to show you how
they work. Fun. Any one game will be worth
more than the price of the book for the fun
you and your family will have with it. $7.50.*
• SCELBI'S FIRST BOOK OF COMPUTER
GAMES Need a game for your 8008 or 8080
microprocessor? Has three popular games,
"Space Capture," "Hexpawn," and "Hang-
man." Complete flowcharts, logic description,
program listing, and instructions are provided.
A must for the game freak! $14.95.*
Use the order card in the back of this magazine or itemize your order on a separate piece of paper and mail to:
Kilobaud Book Department Peterborough NH 03458
Be sure to include check or detailed credit card information
% Add $1.00 shipping & handling charge for each order. Note: Prices subject to change on books not published by 73 Magazine.
software* programming jiii
• DISCOVERING BASIC - A Problem
Solving Approach by Robert E. Smith deals
with progressively more complex problems
which allow the reader to discover the vocab-
ulary of BASIC language as he develops skill
and confidence in putting it to work. Clear
and concise explanations. Problems used
cover a wide range of interests — insurance,
geometry, puzzles, economics, etc. $6.85.*
• BASIC New 2nd Edition, by Bob Albrecht.
Self-teaching guide to the computer language
you will need to know for use with your
microcomputer. This is one of the easiest
ways to learn computer programming.
$4.95.*
• A QUICK LOOK AT BASIC by Donald D.
Spencer. A perfect reference for the beginning
programmer. Assumes that the reader has no
previous programming experience and is a
self-teaching guide for the individual desiring
to learn the fundamentals of BASIC. $4.95.*
• MY COMPUTER LIKES ME ...WHEN I
SPEAK BASIC An introduction to BASIC . . .
simple enough for your kids. If you want to
teach BASIC to anyone quickly, this book is
the way to go. $2.00.*
• FUN WITH COMPUTERS AND BASIC by
Donald D. Spencer, contains an easy-to-under-
stand explanation of the BASIC Programming
Language and is intended for persons who
have had no previous exposure to computer
programming. Over half the book is devoted
to problems using games, puzzles, and math-
ematical recreations. A superior book for
self-teaching and learning computer pro-
gramming. $6.95.*
• SIXTY CHALLENGING PROBLEMS
WITH BASIC SOLUTIONS by Donald Spen-
cer, provides the serious student of BASIC
programming with interesting problems and
solutions. No knowledge of math above
algebra required. Incudes a number of game
programs, as well as programs for financial
interest, conversions and numeric manipula-
tions. $6.95.*
• THE SECRET GUIDE TO COMPUTERS
Parts 1, 2, and 3 by Russ Walter. Part One
describes computers in general, and after
reading for ten minutes you will be writing
simple BASIC programs! Part Two discusses
computer applications. It's one thing to
master the syntax of a language such as
BASIC and another to solve problems using
the new tool. Part Three describes program-
ming languages. Ever heard of APL and
QLISP? BASIC is not the only language used
to program computers. 7th Edition. Part I -
$2.75*; Part II - $2.50*; Part III - $3.50.*
• SOME COMMON BASIC PROGRAMS
published by Adam Osborne & Associates,
Inc. Perfect for non-technical computerists
requiring ready-to-use programs. Business pro-
grams, plus miscellaneous programs. Invalu-
able for the user who is not an experienced
programmer. All will operate in the stand-
alone mode. $7.50 paperback.*
• Scientific Research Instruments' BASIC
SOFTWARE LIBRARY is a complete do-it-
yourself kit. Written in everybody's BASIC
immediately executable in ANY computer
with at least 4K, no other peripherals needed.
Vol. I contains business and recreational
programs and is 300 pages. Vol. II is 260
pages and contains math, engineering, sta-
Fun
with
|iputers
I and
BASIC
aid D. Spencer
t
THE SECRET
BASIC
SOFT WAR
LI BR
FORTRAN
Camelot
tisticsand plotting programs. Vol. Ill contains
money managing, advanced business programs
such as billing, A/R, inventory, payroll, etc.
Vol. IV contains general purpose programs
like loans, rates, retirement, plus games:
Poker, Enterprise (take charge while Capt.
Kirk is away). Football and more! Vol. V is
filled with experimenter's programs including
games, pictures and misc. problems like
"logic." Vols. I & II $24.95,* Vol. Ill
$39.95,* Vol. IV & V $9.95 each.*
• MICROPROCESSOR PROGRAMMING for
Computer Hobbyists by Neill Graham is for
the hobbyist interested in intermediate and
advanced techniques of programming and
data structuring. Written to take up where the
computer manufacturers' instruction manuals
and the introductory programming language
texts leave off. $8.95.*
• ADVANCED BASIC — Applications and
Problems by James Coan is for those who
want to extend their expertise with BASIC.
Offers advanced techniques and applications.
$6.95.*
• 8080 PROGRAMMING FOR LOGIC
DESIGN Ideal reference for an in-depth
understanding of the 8080 processor. Appli-
cation-oriented and the 8080 is discussed in
light of replacing conventional, hard-wired
logic. Practical design considerations are pro-
vided for the implementation of an 8080-
based control system. $7.50.*
• 8080 SOFTWARE GOURMET GUIDE
AND COOKBOOK If you have been spending
too much time developing simple routines for
your 8080, try this new book by Scelbi
Computing and Robert Findley. Describes
sorting, searching, and many other routines
for the 8080 user. $9.95.*
• 6800 PROGRAMMING FOR LOGIC
DESIGN Oriented toward the industrial user,
this book describes the process by which
conventional logic can be replaced by a 6800
microprocessor. Provides practical informa-
tion that allows an experimenter to design a
complete micro control system for the
"ground up." $7.50.*
• 6800 SOFTWARE GOURMET GUIDE &
COOKBOOK If you have been spending too
much time developing routines for your 6800
microprocessor, try the new book by Scelbi
Computing and Robert Findley. Describes
sorting, searching, and many other routines
for the 6800 user. $9.95.*
• FORTRAN PROGRAMMING by Donald
Spencer. FORTRAN was designed for com-
plex numeric calculations; and possesses ex-
tended I/O capability. It is easily learned, as it
is an English-like computer language. $7.50.*
• FORTRAN WORKBOOK by Donald
Spencer. Provides practical examples and
problems to solve. Flowcharting is also dis-
cussed. Almost all micros support BASIC — it
won't be long before FORTRAN is common-
place. $3.95.*
•CHEMISTRY WITH A COMPUTER by Paul
A. Cauchon, contains a collection of tutorial,
simulation and problem-generation computer
programs. Usable with almost any chemistry
course in the high school or college level.
$9.95.*
Use the order card in the back of this magazine or itemize your order on a separate piece of paper and mail to:
Kilobaud Book Department Peterborough NH 03458
Be sure to include check or detailed credit card information
Add $1.00 shipping & handling charge for each order. Note: Prices subject to change on books not published by 73 Magazine.
hardware ittiii
• MICROPROCESSOR INTERFACING
TECHNIQUES by Austin Lesea & Rodnay
Zaks will teach you how to interconnect a
complete system and interface it to all the
usual peripherals. It covers hardware and
software skills and techniques, including the
use and design of model buses such as the
IEEE 488 or $100. $9.95.*
•TTL COOKBOOK by Donald Lancaster.
Explains what TTL is, how it works, and how
to use it. Discusses practical applications, such
as a digital counter and display system, events
counter, electronic stopwatch, digital volt-
meter, and a digital tachometer. $8.95.*
• CMOS COOKBOOK by Don Lancaster.
Details the application of CMOS, the low
power logic family suitable for most appli-
cations presently dominated by TTL. Re-
quired reading for every serious digital
experimenter! $9.95.*
• TVT COOKBOOK by Don Lancaster,
describes the use of a standard television
receiver as a microprocessor CRT terminal.
Explains and describes character generation,
cursor control and interface information in
typical, easy-to-und erst and Lancaster style.
$9.95.*
• BUI LD-IT BOOK OF DIGITAL ELEC-
TRONIC TIMEPIECES by Robert Haviland is
a data-packed guide to building every time-
keeping device you can imagine: rugged ship-
board clocks, second-splitting digital 1C
chronometers, decorator digital clocks, a pre-
cision timer, a frequency-period meter, a tide
and moon clock, an automatic alarm setter,
etc. Including full-size printed circuit board
layouts. $6.95.*
general
• THE "COMPULATOR" BOOK - Building
Super Calculators & Minicomputer Hardware
with Calculator Chips by R. P. Haviland,
provides ideas, design info and printed circuit
boards for calculator chip projects, measure
time, tie in with a Teletype to create a
virtually infinite memory system, and count-
less other functions. $7.95.*
TEST EQUIPMENT LIBRARY
•VOL I COMPONENT TESTERS Build your
own test equipment and save a bundle (and
have a lot of fun). Volume I of the 73 Test
Equipment Library shows you how to build
and use transistor testers (8 of 'em), three
diodes testers, 3 1C testers, 9 voltmeters and
VTVMs, 8 ohmmeter, 3 inductance meters,
and a raft of other gadgets for checking
temperature, crystals, Q, etc. $4.95.*
• VOL II AUDIO FREQUENCY TESTERS If
you're into audio such as digital cassette
recording, RTTY, Baudot vs ASCII, SSTV,
SSB, Touchtone or even hi-fi you'll want to
have this book full of home built test equip-
ment projects. Volume II $4.95.*
• VOL III RADIO FREQUENCY TESTERS
This is of more interest to hams and CBers:
test equipment you can build for checking
out transmitters and receivers, signal gener-
ators, noise generators, crystal calibrators,
GDOs, dummy loads. $4.95.*
• VOL. IV 1C TEST EQUIPMENT Become a
trouble-shooting wizard. All you need to
know about pulse, audio and sync generators,
frequency counters, digital component
testers, logic probes and more! Plus a cumu-
lative index for all four volumes of the 73
Test Equipment Library. $4.95.*
• COMPUTER DICTIONARY by Donald D.
Spencer. Defines words and acronyms used by
computerists in a clear, easy to understand
style. This reference is a must for the
individual getting started in the world of
microcomputers. $5.95.*
• BRAND NEW DICTIONARY This micro-
computer dictionary fills the need to become
quickly acquainted with the terminology and
nomenclature of the revolution in computers.
There is also a comprehensive electronics/
computer abbreviations and acronyms
section. $1 5.95.*
• THE UNDERGROUND BUYING GUIDE
Here is a handy guide for the electronics
enthusiast. Over 600 sources of equipment
and literature are provided. Cross-referenced
for ease of use. Electronic publishing houses
are'also listed. $5.95 each.*
Period^ Gu^c
* ,u
MICROCOMPUTER
• 1976 PERIODICAL GUIDE FOR COM-
PUTERISTS is a 20-page book which indexes
over 1,000 personal computing articles for the
entire year of 1976 from Byte, Creative
Computing, Digital Design, Dr. Dobbs
Journal, EDN, Electronic Design, Electronics,
Interface Age, Microtrek, Peoples Computer
Company, Popular Electronics, QST, Radio
Electronics, SCCS Interface and 73 Amateur
Radio. Price $3.00.* New January — June
1977 Edition (includes Kilobaud) — $3.00.*
• TYCHON'S 8080 OCTAL CODE CARD
Slide rule-like aid for programming and
debugging 8080 software contains all the
mnemonics and corresponding octal codes.
Also available, Tychon's 8080 Hex Code
Card, same as above only has hex codes
instead of octal. $3.00 each.*
amateur radio books
• NOVICE STUDY GUIDE The most
complete Novice study guide available. It is
brand new. This is not only invaluable for
anyone wanting to get started in amateur
radio, but also it is about the only really
simple book on the fundamentals of elec-
tricity and electronics. $4.95.*
• GENERAL CLASS STUDY GUIDE Takes
over on theory where the Novice book leaves
off. You'll need to know the electronic
theory in this to work with computers and
you'll not find an easier place to get the
information. $5.95.*
• SSTV HANDBOOK This excellent book
tells all about it, from its history and basics to
the present state-of-the-art techniques. Hard-
bound $7,* Softbound $5.*
• VHF ANTENNA HANDBOOK This new
handbook details the theory, design and
construction of hundreds of different VHF
and UHF antennas. Packed with fabulous
antenna projects you can build. $4.95.*
• WEATHER SATELLITE HANDBOOK
Simple equipment and methods for getting
good pictures from the weather satellite. Dr.
Taggart WB8DQT $4.95.*
• THE NEW RTTY HANDBOOK is a brand
new 1977 edition and the only up-to-date
RTTY book available. The state-of-the-art has
been changing radically and has made all
previous RTTY books obsolete. It has the
latest circuits, great for the newcomer and the
expert alike. $5.95.*
Use the order card in the back of this magazine or itemize your order on a separate piece of paper and mail to:
Kilobaud Book Department Peterborough NH 03458 Be sure to include check or detailed credit card information
^fc Add $1.00 shipping & handling charge for each order. Note: Prices subject to change on books not published by 73 Magazine.
new ADDITIONS TO TW€
KB BOOK HOOK
• MICROPROCESSORS FROM CHIPS TO
SYSTEMS by Rodnay Zaks is a complete and
detailed introduction to microprocessors and
microcomputer systems. No preliminary
knowledge of computers or microprocessors is
required to read this book, although a basic
engineering knowledge is naturally an
advantage. Intended for all wishing to under-
stand the concepts, techniques and com-
ponents of microprocessors in a short time.
$9.95.*
• YOUR OWN COMPUTER by M. Waite and
M. Pardee. The personal computer has been
touted as the next consumer product. But
most individuals still wonder why. Much
technical material has been written but there
is little material for the average individual
without an extensive background in elec-
tronics. This book removes the stigma of
complexity that surrounds the computer and
has succeeded in providing a simple easy-to-
understand guide to these units. $1.95.*
• BEGINNING BASIC by Dr. Paul Chirlian
provides a really basic BASIC that covers all
the topics in simple, easy-to-understand
language. Nothing is left out, everything is
presented in clear, step-by-step fashion. This
book will make a good BASIC programmer of
any reader. $9.95.*
• THE COMPUTER QUIZ BOOK by Donald
D. Spencer is written for readers who would
like to test themselves on basic computer
concepts. It may be used effectively by
students, teachers, laymen, programmers, per-
sonal computer users, or anyone else inter-
ested in checking their knowledge of
computer concepts. $5.95.*
• PAYROLL WITH COST ACCOUNTING -
IN BASIC by L. Poole & M. Borchers includes
program listings with remarks, descriptions,
discussion of the principle behind each pro-
gram, file layouts, and a complete user's
manual with step-by-step instructions, flow
charts, and simple reports and CRT displays.
Payroll and dost accounting features include
separate payrolls for up to 10 companies,
time-tested interactive data entry, easy correc-
tion of data entry errors, job costing (labor
distribution), check printing with full deduc-
tion and pay detail, and 16 different printed
reports, including W-2 and 941. Price only
$1250.*
• HOW TO BUY & USE MINICOMPUTERS
AND MICROCOMPUTERS by Wm. Barden,
Jr. This book discusses these smaller brethren
of computers and shows how the reader can
become a part of the revolution — how he can
own and use a functioning computer system
in his home to do a variety of practical or
recreational tasks. $9.95.*
• THE 8080A BUGBOOK-MICROCOM-
PUTER INTERFACING AND PROGRAM-
MING is written for the 8080 user who has a
knowledge of digital elements and operations.
This book will be invaluable, as it explains the
fundamental tasks of microcomputer inter-
facing and the associated microcomputer I/O
programming for 8080-based microcom-
puters. Only $9.95.*
• HOW TO PROGRAM MICROCOMPUTERS
by Wm. Barden, Jr. Here is a guide to
assembly language programming of the Intel
8080, Motorola MC6800, and MOS Tech-
nology MCS6502 microprocessors. It is
written especially for beginning programmers
with hobbyist microcomputers based on one
of these three chips. The topics covered range
from data manipulations at the bit level up to
data handling of tables and lists, and from
simple adds and subtracts up to floating-point
operations. $8.95.*
• TAKE A CHANCE WITH YOUR CAL-
CULATOR Lennart Rade wrote this book to
help you to discover the world of probability
with your programmable calculator. You will
need NO previous experience either in prob-
ability theory or in programming to learn
both from this book. $8.95.*
• A STEP BY STEP INTRODUCTION TO
8080 MICROPROCESSOR SYSTEMS by
D.L Cohn & J.L. Melsa. This microprocessor
book is written for people who don't know
anything about microprocessors but who wish
they did. The step-by-step presentation does
not require any computer or electronics back-
ground, so anyone who is interested can
follow it. However, the book is not only
intended for beginners. Engineers and tech-
nicians who are familiar with electronics will
find the software descriptions valuable in
updating their skills. Computer professionals
will find the detailed treatment of the 8080
architecture and instruction set useful. Price
$7.95.*
Use the order card in the back of this magazine or itemize your order on a separate piece of paper and mail to:
Kilobaud Book Department Peterborough NH 03458
Be sure to include check or detailed credit card information
% Add $1.00 shipping & handling charge for each order. Note: Prices subject to change on books not published by 73 Magazine.
F-8 MICROCOMPUTER ON SINGLE BOARD
WITH F-8 CPU, FAIRBUG PSU, 3853 SMI-DEVELOPED FOR DESIGN ENGINEERS AND SERIOUS HOBBYISTS
F-8 MICROCOMPUTER Model 1080
Design Features Include:
• Buffered address and data bus to S-
100 expansion connector • 4K program
memory - 2708 sockets • 2K onboard
page selectable static RAM • FAIRBUG
monitor plus IK 2708 custom monitor •
Independent FAIRBUG scratch pad RAM
• Memory protect switches • Buffered
high speed data input port • l/O/interface
selection (all 4 ports available) • Two
sockets for I/O expansion • 20 ma or RS
232 interface • Quality solder masked
PC board • Program control timers •
Documentation • 64K addressing range
• Prime parts
.1249
00
Price postpaid Assembled $299
California lesidents add 6% tax
Send check, money order, corporate
purchase order or COD orders to:
19824 Ventura Blvd.
Woodland Hills, CA 91 364
(213) 340-8843
C62
143
1977 <K<B BACK ISSUES
kilobaud
The Computer Hobbyist Magazine
kilobaud
The Small Computer Magazine
kilobaud
The Small Computer Magazine
kilobaud
The Small Computer Magazine
kilobaud
The Small Computer Magazine
kilobaud
I he Small Computet Magazine
kilobaud
The Small Computer Magazine
kilobaud
The Small Computer Magazine
kilobaud
The Small Computer Magazine
kilobaud
The Small Computer Magazine
kilobaud
The Small Computer Magazine
kilobaud
The Small Computer Magazine
Issue #1— $5.00; #2 thru
Please send Kilobaud ba
issues JAN 77-$5
issues FEB 77— $3
issues MAR 77— $3
issues APR 77— $3
2— $3.00 each.
( issues:
issues MAY 77-$3
issues JUNE 77— $3
issues JUL 77-$3
issues AUG 77— $3
issues SEPT 77-$3
issues OCT 77— $3
issues NOV 77 -$3
issues DEC 77— $3
Add $1.00 shipping & handling for each order.
□ BankAmericard/Visa □ Master Charge □ American Express
Card # Expiration date
Signature
TOTAL:.
Name
Have credit card # handy and call
TOLL FREE (800) 258-5473
5/78 Kilobaud • Peterborough NH 03458
Address
City State Zip
144
AiUUJUJULfiJLO^
MAILING LIST
Completely re-written and updated to include dealers, clubs,
publications, and manufacturers. (It's the one we use for our
mailings and we update it daily.) The list has over 2000
names painstakingly gathered from manufacturers, magazine
ads, new product releases, hobby computer shows, and direct
mail. You can buy this list printed on self-sticking labels for
only $50.
kilobaud
pETEnboRouqh Nh 05458
4
So
©
o
©
©
o
o
©
©
o
o
o
©
o
o
o
o
o
o
o
o
o
o
o
o
©
o
©
o
©
o
o
o
o
o
©
©
©
SEND FOR THESE GREAT BOOKS TODAY!
• HOBBY COMPUTERS ARE HERE If you (or a friend)
want to come up to speed on how computers work
. . . hardware and software . . . this is an excellent book.
It starts with the fundamentals and explains the cir-
cuits, the basics of programming, along with a couple of
TVT construction projects, ASCII-Baudot, etc. This book
has the highest recommendations as a teaching aid for
newcomers. $4.95
• THE NEW HOBBY COMPUTERS! This book takes it
from where “Hobby Computers Are Here” leaves off,
with chapters on Large Scale Integration, how to choose
a microprocessor chip, an introduction to programming,
low cost I/O for a computer, computer arithmetic, check-
ing memory boards, a Baudot monitor/editor system, an
audible logic probe for finding those tough problems, a
ham’s computer, a computer QSO machine . . . and
much, much more! Everything of interest is there in one
volume. Don’t miss this tremendous value! Only $4.95
Have your credit card handy and call toll-free
800-258-5473, or use the card in the back of this
magazine and send your check to:
KILOBAUD Book Dept. • Peterborough, NH 03458
(Be sure to include $1.00 per order for shipping & handling.)
o
o
o
o
o
o
o
1977 kilobaud Binders
Keep your Kilobauds from being lost or
damaged in these handsomely appointed binders
with rich blue covers and gold lettering. Each
binder holds 12 issues, making an EXCELLENT
REFERENCE HANDBOOK. Several binders
form a quality library you can be proud of.
Order yours today . . . only $6.50 each ... 2 for
$ 12 . 00 .
Use the order card in the back of this magazine or itemize your
order on a separate piece of paper and mail to:
Kilobaud Binders Peterborough NH 03458
Be sure to include check or detailed credit card information
Add $1.00 shipping & handling charge for each order.
MOVING?^
Let us know 8 weeks in advance so that you won’t miss
a single issue of Kilobaud.
Attach old label where indicated and print new address
in space provided. Also include your mailing label
whenever you write concerning your subscription. It
helps us serve you promptly. Write to:
• KILOBAUD • PETERBOROUGH NH 03458 •
□ Address change only □ Payment enclosed
□ Extend subscription (1 extra BONUS issue)
□ Enter new subscription □ Bill me later
□ 1 year $15.00
If you have no label handy , print OLD address here.
cq Name Call
^ Address
fcfccity State Zip
print NEW address here:
Name Call
Address
State-
Zip
145
it's a good bet the company
you bought your computer from
doesn't even make peripherals!
Its no great surprise! Most
computer companies got their
start in the digital logic end of
the business. They were great
at building calculators and
later computers but when it
came right down to it, most
just didn't have the experi-
ence necessary to build the
peripherals to support their
computer products. And that
left a vacuum!
At Heath we had the advan-
tage. Our years of experience
in electronic kit design gave
us plenty of background with
not only digital logic but
mechanical and video design
as well. And our assembly
manuals and documentation
are world-famous for easy to
understand instructions.
We built the world s first digi-
tal color television, a unique
fully synthesized FM tuner,
digital frequency counters,
clocks -even a digital bath-
room scale.
So when we entered the per-
sonal computing market we
had the "know-how" to build
not only our outstanding H8
and Hll, 8 and 16 -bit comput-
ers, but, in addition, a com-
plete line of supporting
peripheral kits !
Select the H9 Video Terminal,
the H10 Papertape Reader/
Punch, and very soon our
own, complete, Floppy Disk
system. Each was designed
with the systems approach in
mind. Each was conceived to
integrally mesh with not only
our own computers, but
through our set of sophisti-
cated interfaces, most others
as well. And in that Way we re
making every effort to fill the
vacuum the others left!
So when you're ready to
communicate with your com-
puter turn to Heath. We've got
the peripheral kits you'll need
and at prices you can afford.
Maybe the company who sold
you your computer didn't
think about peripherals - but
we sure did! And come to
think about it maybe that's
why you should come to
Heath. . .in the first place.
Heathkit
Heathkit Catalog
Read about nearly
400 money-saving,
fun-to-build
. electronic kits.
\ Use coupon to send for
: \your mail order catalog
;\ or bring coupon to a
2? Heathkit Electronic
Center for your catalog.
r
_ HEATH
Schlumberger
Heath Company Dept. 351-410
Benton Harbor, Ml 49022
Please send me my FREE Catalog. I am not on your mailing list.
L
Name
Address.
City
CP-146
State.
Zip.
J
AVAILABLE LOCALLY AT
HEATHKIT ELECTRONIC CENTERS
(Units of Schlumberger Products
Corporation) Retail prices on some
products may be slightly higher.
ARIZONA: Phoenix, 85017, 2727 W. Indian School
Rd., Phone: 602-279-6247; CALIFORNIA: Anaheim,
92805, 330 E. Ball Rd., Phone; 714-776-9420; El
Cerrito, 94530, 6000 Potrero Ave., Phone: 415-236-
8870; Los Angeles, 90007, 2309 S. Flower St.,
Phone: 213-749-0261; Pomona, 91767, 1555 Orange
Grove Ave. N., Phone: 714-623-3543; Redwood
City, 94063, 2001 Middlefieid Rd., Phone: 415-365-
8155; Sacramento, 95825, 1860 Fulton Ave., Phone:
916-486-1575; San Diego (La Mesa, 92041), 8363
Center Dr., Phone; 714-461-0110; San Jose (Camp-
bell, 95008), 2350 S. Bascom Ave., Phone: 408-
377-8920; Woodland Hills, 91364, 22504 Ventura
Blvd., Phone: 213-883-0531; COLORADO: Denver,
80212, 5940 W. 38th Ave., Phone: 303-422-3408;
CONNECTICUT: Hartford (Avon, 06001), 395 W.
Main St. (Rte. 44), Phone: 203-678-0323; FLORIDA:
Miami (Hialeah, 33012), 4705 W. 16th Ave., Phone:
305-823-2280; Tampa, 33614, 4019 West Hills-
borough Ave., Phone: 813-886-2541; GEORGIA:
Atlanta, 30342, 5285 Roswell Rd., Phone: 404-252-
4341; ILLINOIS: Chicago, 60645, 3462-66 W. De-
von Ave., Phone: 312-583-3920; Chicago (Downers
Grove, 60515), 224 Ogden Ave., Phone: 312-852-
1304; INDIANA: Indianapolis, 46220, 2112 E. 62nd
St, Phone: 317-257-4321; KANSAS: Kansas City
(Mission, 66202), 5960 Lamar Ave., Phone: 913-
362-4486; KENTUCKY: Louisville, 40243, 12401
Sheibyville Rd., Phone: 502-245-7811; LOUISIANA:
New Orleans (Kenner, 70062), 1900 Veterans
Memorial Hwy., Phone: 504-722-6321; MARYLAND:
Baltimore, 21234, 1713 E. Joppa Rd., Phone: 301-
661-4446; Rockville, 20852, 5542 Nicholson Lane,
Phone: 301-881-5420; MASSACHUSETTS: Boston
(Peabody, 01960), 242 Andover St., Phone: 617-
531-9330; Boston (Wellesley, 02181), 165 Wor-
cester Ave. (Rt. 9 just west of Rt. 128), Phone:
617-237-1510; MICHIGAN: Detroit, 48219, 18645
W. Eight Mile Rd., Phone: 313-535-6480; E, De-
troit, 48021, 18149 E. Eight Mile Rd., Phone: 313-
772-0416; MINNESOTA: Minneapolis (Hopkins,
55343), 101 Shady Oak Rd., Phone: 612-938-6371;
MISSOURI: St Louis (Bridgeton), 63044, 3794
McKelvey Rd., Phone: 314-291-1850; NEBRASKA:
Omaha, 68134, 9207 Maple St., Phone: 402-391-
2071; NEW JERSEY: Fair Lawn, 07410, 35-07
Broadway (Rte. 4), Phone: 201-791-6935; Ocean,
07712, 1013 State Hwy. 35, Phone: 201-775-1231;
NEW YORK: Buffalo (Amherst, 14226), 3476 Sheri-
dan Dr., Phone: 716-835-3090; Jericho, Long Is-
land, 11753, 15 Jericho Turnpike, Phone: 516-334-
8181; Rochester, 14623. 937 Jefferson Rd., Phone:
716-244-5470; White Plains (North White Plains,
10603), 7 Reservoir Rd. t Phone: 914-761-7690;
OHIO: Cincinnati (Woodlawn, 45215), 10133
Springfield Pike, Phone: 513-771-8850; Cleveland,
44129, 5444 Pearl Rd., Phone: 216-886-2590; Col-
umbus, 43229, 2500 Morse Rd., Phone: 614-475-
7200; Toledo, 43615, 48 S. Byrne Rd., Phone: 419-
537-1887; PENNSYLVANIA: Philadelphia, 19149,
6318 Roosevelt Blvd., Phone: 215-288-0180; Frazer
(Chester Co.), 19355, 630 Lancaster Pike (Rt. 30),
Phone: 215-647-5555; Pittsburgh, 15235, 3482 Wm.
Penn Hwy., Phone: 412-824-3564; RHODE ISLAND:
Providence (Warwick, 02886), 558 Greenwich
Ave., Phone: 401-738-5150; TEXAS: Dallas, 75201,
2715 Ross Ave., Phone: 214-826-4053; Houston,
77027, 3705 Westheimer, Phone: 713-623-2090;
VIRGINIA: Alexandria, 22303, 6201 Richmond
Hwy., Phone: 703-765-5515; Norfolk (Virginia
Beach, 23455), 1055 Independence Blvd., Phone:
804-460-0997; WASHINGTON: Seattle, 98121, 2221
Third Ave., Phone. 200-082-2172; WISCONSIN;
Milwaukee, 53216, 5215 W. Fond du Lac, Phone:
414-873-8250.
146
This 8-bil machine,
by itself, is as versatile
as a lot of systems
that include peripherals
Be sure to use <
coupon on
facing page of this magazine
to order your FREE
Heathkit Catalog!
§JM
-Sill
jjjj
- -
gl ii§ ^jjj
..
-
P*-™” |
!
Skeptical? For starters, because of its
unique design
the H8 is the
only machine
in its price
class that
offers full
system inte-
gration, yet,
with just 4K of
memory and
using only
its "intelligent"
front panel for I/O, may be operated
completely without peripherals!
In addition, by using the features of
its built-in Pam-8 ROM panel
control program, the H8 actually
allows you to dig in and examine
machine level circuitry
Responding to simple instructions
the "intelligent" panel displays
memory and register contents, lets
you inspect and alter them even
during operation. And for greater
understanding, the front panel
permits you to execute programs a
single instruction at a time. The
result is a powerful, flexible learn-
ing tool that actually lets you "see"
and confirm each detail of H8's
inner workings.
If you need further evidence, con-
sider the fact that H8's system
orientation allows you an almost
unlimited opportunity for growth.
Memory is fully expandable, the
8080A CPU extremely versatile,
and with the addition of high speed
serial and parallel interfacing
you gain the added flexibility of
I/O operation with tape, CRT con-
soles, paper tape reader/punches,
and soon floppy disk systems!
The H8 offers superior documen-
tation including complete step-by-
step assembly and operation
manuals, is backed by 54 years of
Heath reliability, and comes
complete with BASIC, assembler,
editor, and debug software —
others charge over $60 for!
H8, simplicity for the
System Engineered
for Personal Computing
J jJJfEB! One-Stop
Component Center
NEW!! EXPANDED PRODUCT LINE. . .
7400
7402
7404
7408
7410
7420
7430
7432
7442
7490
7492
7493
74107
74121
74122
74123
74125 (DM8093)
74154
74161 (DM9316)
741 76 (DM8280)
74177 (DM8281)
74192
74193
74367 (DM8097)
Low Power TTL Schottky
74LSOO 74LS73 74LS109
74LS02 74LS74 74LS123
74LS04 74LS75 74 LSI 36
74LS08 74LS83 74 LSI 38
74LS10 74LS85 74LS151
74LS20 74LS86 74LS17S
74LS30 74LS90 74LS367
74LS32
LM311V
LM320K 5 ( 7905)
LM320T 5 (7905)
LM320T 12(7912)
LM320T 15(7915)
LM339N
LM340T 6 (78051
LM340T 12(7812)
LM340T 15(7815)
LM555V
LM556N
LM567V
LM723N
LM739N
LM741V
LM747N
LM1458V (5558V)
LM1488N
LM1489N
LM 1556V
XR2206
CA3080
CA3130
CA3140
LM390ON (CA3401I
CAPACITORS
Aluminum Electrolytic
Imld 50V 100m «d 60V
4.7mfd 50V 220mfd 50V
lOmfd 50V 470mfd 50V
22m fd 50V lOOOmfd 25V
47mld 50V 2200mid 16V
Ceramic Disc
.001m td 50V
,0O47m(d 50V
■ Olmfd 60V
022mfd 50V
047mfd 50V
. Imfd 50V
MICROPROCESSOR
10p( 50V
47pf 50V
100p( 50V
220p» 50V
330pf 60V
470p1 50V
8080 A
8212
8224
8228
C/MOS
4000 4020 4050
4001 4023 4051
4010 4024 4069
4011 4029 4071
Display LEDS Discrete
MAN2 XC556 Rad
DL704 XC556 G raan
DL707 XC656-Y«llow
DL747 CLIPUTCS-Rad.
DL 750 Graan, YallOw
2101
2102
21L02
7489
MM5262
6800 1 702 A
6810 82S23
6830L8 2708
AY-5 1013 DM8835N
2513/2140 N8T97
MM5314I Clock Chip)
SOCKETS
8 pin low profile 14 pin ware w
14 pin low profile 16 pin wire w
16 pin low profile 24 pm wire w
24 pin low profile 40 pin wire w
40 pin low profile TO-3 Socket
14 pin plug TO 5 Socket
16 pin plug Molex Pint
Dipped Tantalum
d 36V
,22mfd 35V
.33mfd 35V
.47mfd 35V
68mfd 35V
Imfd 35V
1.5mfd 35V
2. 2mfd 26V
3.3mfd 25V
4.7mfd 25 V
6.8mfd 25V
lOmfd 25V
15mfd 25V
33mfd 25V
Polyester Mylar
.OOlmfd 100V ,022mfd 100 V
OOlSmtd 100V ,047mfd 100V
0022mfd 100V .Imfd 1O0V
0047mfd 100V .22mfd 100V
.Olmfd 10OV
CRYSTALS
CY 1 A CY2A CY 12A
TEST CLIPS
14 pin clip 16 pin clip
FUSE HOLDERS
HKP-3AG
P.C. POTENTIOMETERS
Single-Turn 15-Turn
84 OP IK 830P1K
840P5K 830P5K
84 OP 1 OK 83OP10K
840P50K B3OP50K
840P100K 83OP10OK
840P1meg 830P1meg
DIODES
IN751 IN41480N914)
IN4733 IN4001
IN4734 I N 4 004
IN4742 IN4007
IN4744 M0A98O3
TRANSISTORS
Cl 068 1 2N30S5
2N2222A 2N3904
2N2907A 2N3906
MJE2955 2NS129
MJE30SS 2N5139
CONNECTORS
DB25P Plug DB25S Socket
DATA BOOKS*
7400/74LS Date Book
CMOS/Linear Data Book
Microprocenor/LED Oata Book
•JIM PAX Product, onlv
HEAT SINKS
Ye W RESISTOR ASST.
X Watt 5%
50 Pc Awortment.
A new component product line of over 260 items developed for the independent dealer. Guaranteed nationally advertised
products. Complete JIM-IPAK program includes national advertising, direct mail programs, store display racks, stock
rotation plan and return policy. For dealer information, write or call JIM-PAK, 1021 Howard Avenue, San Carlos, CA
94070 (415)592-8097.
SEE JIM-PAK AT NEWCOM '78 - MAY 2, 3 and 4, 1978 AT THE LAS VEGAS HILTON
[★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★★|
SEE YOUR LOCAL 5 DEALER TODAY...
ALABAMA
ILLINOIS (Continued)
OHIO
Mobile
Lafayette Radio Electronics
Oak Park
Spectronics
Bucyrus
Mead Electronics
ARIZONA
Schaumburg
Data Domain
j* Cincinnati
Digital Design
Fountain Hills
P & C Communications
INDIANA
Qeyton
Altair Computer Center
Tempe
Computerworld Inc.
East Chicago
Aero Electronics
Reynoldsburg
Universal Amateur Radio
Yuma
Yuma Electronics
Hammond
Quantum Computer Works
Steubenville
Hosfelt Electronics
CALIFORNIA
IOWA
OKLAHOMA
Bellflower
Earl's Hobby Shop
Indianola
Electronix Limited
Suymon
Sound Service
Berkeley
At Lasher Electronics
KANSAS
♦ Oklahoma City
Bits. Bytes & Micros
Fontana
Fontana Electronics
Wichita Amateur Radio Equipment Co.
OREGON
Fullerton
Orvac Electronics Inc.
KENTUCKY
Qggverton
Altair Computer Center
Long Beach
Scott Radio Supply
Lexington Radio-Electronic Equipment Co.
Coos Bay
Herrick Electronics
Mission Viejo
Tower Electronics Corp.
LOUISIANA
Ontario
Miller Electronics
Monterey
Zackit
Baton Rouge
Davis Electronics Supply
Salem
Computer Pathways
Oceanside
Electronic Center
MARYLAND
PANAMA
Palo Alto
Zack Electronics
Baltimore Computer Workshop of Baltimore
Panama City
Son i tel. S.A.
Pasadena
Dow Radio Inc.
Baltimore
Everything Electronic
PENNSYLVANIA
Sacramento
The Radio Place
Lavale
J & M Electronics
Hershey
Microcomputer Systems Inc.
Sacramento
Zackit
Rockville
Computer Workshop
Murraysville
Computer Workshop
San Carlos
J & H Outlet Store
Towson
Baynesville Electronics
of Pittsburgh
San Diego Radio Shack A.S.C. Mira Mesa
Towson
Computers. Etc.
RHODE ISLAND
San Fernando
San Fernando Electronics
MASSACHUSETTS
Cranston
Jabbour Electronics City
San Francisco
Zack Electronics
Medford
Tufts Electronics
Pawtucket
Jabbour Electronics City
San Jose
Quement Electronics
North Adams
Electronics Supply Center
SINGAPORE
Intertrade (PTE) Ltd.
San Luis Obispo
Mid-State Electronics
Waltham
Computer Mart Inc.
SOUTH CAROLINA
Santa Monica
Mission Control
MICHIGAN
No. Charleston
Technical Services Inc.
Sunnyvale
Sunnyvale Electronics
Grand Rapids
Micro Computer World
TENNESSEE
Vallejo
Zackit
Lansing
Fulton Radio Supply
Clarksville
Masstronics
Walnut Creek
Byte Shop Computer Store
Mt. Clemens
The Computer Store
Knoxville
Byte Shop
CANADA
MINNESOTA
Memphis Sere-Rose & Spencer Electronics
Alberta (Calgary)
The Computer Shop
Duluth
Northwest Radio of Duluth
Oak Ridge
Computer Denn
Ontario (Willowdale ) Home Computer Centre
Eagan
Computer Room Inc.
TEXAS
COLORADO
MISSOURI
Dallas
CompuShop
Aurora
Com Co Electronics
El Dorado Springs
Beckman Electronics
Houston
Altair Computer Center
Steamboat Springs
Norm's TV & Electronics
Parkville Computer Workshop of Kansas City
Houston
CompuShop
CONNECTICUT
MONTANA
Houston
Interactive Computers
Bridgeport
Bridgeport Computer
Billings
Conley Radio Supply
San Antonio
Sherman Electronics Supply
FLORIDA
NEBRASKA
VIRGINIA
Ft. Lauderdale
Computers For You
Lincoln
Altair Computer Center
Alexandria
Computer Hardware Store
Lakeland Lakeland Specialty Electronics
Omaha
Omaha Computer Store
Alexandria
Computers Plus
Orlando A Itair Computer Can ter o f Orlando
NEVADA
Charlottesville
Lafayette Electronics
Tampa
AMF Electronics
Las Vegas
Century 23
Richmond
Computers- To-Go
Tampa
Microcomputer Systems
NEW JERSEY
Springfield
Computer Workshop
FRANCE
Bayville A.R.S. Communications Services
of North Virginia
Paris
Computer Boutique
Cherry Hill
Computer Emporium
Virginia Beach
Heathkit Electronics Center
GEORGIA
Hoboken
Hoboken Computer Works
WASHINGTON
Atlanta
Atlanta Computer Mart
Pompton Lakes
Computer Corner
Bellevue
Altair Computer Center
HAWAII
of New Jersey
Longview
Progress Electronics
Aiea
Delcoms Hawaii
Ramsey
Typetronic Computer Store
Pasco
Riverview Electronics
Honolulu
Integrated Circuit Supply
NEW YORK
Seattle
C-Com
IDAHO
Albany
Fort Orange Electronics
Seattle
Empire Electronics
Idaho Falls
Audiotronics
New York
The Computer Stores Inc.
Spokane
Personal Computers
ILLINOIS
New York
Computer Mart of New York
WEST VIRGINIA
Evanston
Tri-State Electronics
Troy
Trojan Electronics
Morgantown
The Computer Corner
Evanston
Itty Bitty Machine Co.
White Plains
The Computer Corner
Morgantown
Electro Distributing Co.
Grove land
Moyer Electronics
NORTH CAROLINA
Mount Prospect
Tri-State Electronics
Durham
Futureworld
Greensboro
Byte Shop
Raleigh
Byte Shop of Raleigh